Clean tech industry primer jefferies (2008)

Jefferies Research

October 2008 Clean Technology Primer

Jefferies & Company, Inc.

Jefferies Research O

ctober 2008C

lean Tech

nolog

y Primer

Key Themes in Clean TechnologyMichael [email protected] (0) 20 7029 8680Laurence Alexander, [email protected] 284 2553Paul Clegg, [email protected] 284 2115

SolarPaul Clegg, CFAMichael McNamaraJames [email protected] (0) 20 7029 8691David [email protected] 284 2175

WindMichael McNamaraJames Harris

Biofuels Laurence Alexander, CFARobin Campbell, [email protected] (0) 20 7029 8678Lucy [email protected] 284 2290

BioplasticsRobin Campbell, Ph.D.Laurence Alexander, CFALucy Watson

WaterAlex Barnett, CFALaurence Alexander, CFALucy Watson

Carbon SequestrationLaurence Alexander, CFAMichael McNamaraPaul Clegg, CFA

Battery TechnologyAlex Barnett, [email protected] 1 5343 6714

Fuel CellsMichael McNamara

Nuclear Debra E. [email protected] 284 2452Laurence Alexander, CFA

Project Finance Laurence Alexander, CFAChris Groobey, Baker & McKenzie [email protected] 835 4240Nathan Read, Baker & McKenzie [email protected] 835 1668

102008 UK Clean Technology Primer_cvr:Layout 1 10/13/2008 9:29 AM Page 1

Member SIPC • © 10/2008 Jefferies & Company, Inc.

US Equity Sales Offices

Atlanta, Georgia 404 264 5000Boston, Massachusetts 617 342 7800Chicago, Illinois 312 750 4700Dallas, Texas 972 701 3000Houston, Texas 713 658 1100Los Angeles, California 310 445 1199New York, New York 212 284 2300San Francisco, California 415 229 1500Short Hills, New Jersey 973 912 2900Stamford, Connecticut 203 708 5900

International Equity Sales Offices

London, England +44 (0) 207 618 3500*Paris, France +33 (0) 1 5343 6700*Tokyo, Japan +81 (0) 3 5226 2411**Zurich, Switzerland +411 227 1600*

*Offices of Jefferies Group, Inc. subsidiaries Jefferies International Limited

**Jefferies Japan Limited

Jefferies US & International Sales Offices

Jefferies & Company, Inc.520 Madison AvenueNew York, NY 10022

www.jefferies.com

Investment Banking

Sales & Trading

Research

Asset Management


Table of Contents

Key Themes in Clean Technology

Solar

Wind

Biofuels

Bioplastics

Water

Carbon Sequestration

Battery Technology

Fuel Cells

Nuclear

Project Finance

Michael McNamara, [email protected], 44 207 029 8680

3

9

47

69

99

115

151

163

173

179

191

Please see important disclosure information on pages 208 - 210 of this report.Page 1 of 212




JIL is Authorised and Regulated by the Financial Services Authority.

EventClean Technology encompasses a wide range of industries andbusiness models that stand to benefit from powerful seculartrends in favor of more efficient use of resources in the face ofrapid demand growth in the emerging economies that is stressingenergy and water supplies. This primer focuses on alternativeenergy (wind, solar, biofuels, etc.), carbon sequestration, greenchemistry, and water. Please refer to the individual reportscontained in this primer for industry-specific details.

Key Points• Industry outlook: Shifting political incentives, credit concerns,

funding pressures, persistent energy price volatility, andconcerns over the pace of capacity additions in several cleantechnology areas are main themes for 2008. Nonetheless,climate change combined with the job creation potential ofrenewable energy as key themes should help offset macrorisks, particularly the impact of volatility in oil prices onsentiment. For solar we expect continued operational strengthalthough we anticipate momentum could stall due to concernsthat abundant polysilicon could translate into a 15-20% declinein module pricing in the coming 12-18 months. For wind weanticipate strong global demand to continue, driven by highenergy prices and supportive government incentive programs.For biofuels, new corn ethanol mandates could allay near-termconcerns over oversupply—but also provide incentives toencourage next-generation biofuels.

• How to play it. In the current environment of extreme marketuncertainty, we suggest that investors focus on "lower-risk"investments characterized by relative certainty of demandand/or proven technologies. We will review some of ourpreferred picks later in this report.

• Financing. Clean Technology is often characterized by highcapital intensity. Examples range from polysilicon plants to windfarms to next generation cellulosic ethanol production. To date,the effects of the credit crunch have not yet taken a significantbite out of clean technology financing although ongoingweakness in the financial sector could impact the pool of fundsavailable while weaker profits could limit the appetite for taxequity investments.

• Carbon sequestration a longer-term theme: An introductoryessay in this report provides an overview of some of the keyfactors driving clean technology equities, assesses some of thecompeting claims on investor capital, and provides an updateon the outlook for carbon sequestration, which we view as oneof the most intuitive, and yet in practical terms more fraught, ofthe clean technologies under consideration.

October 10, 2008

Clean Technology Clean TechnologyKey Themes in Clean Technology

Investment SummaryShifting incentive programs, concerns over the pace of capacityadditions, a macro climate inimical to high-beta names, potentialdifficulty in obtaining financing, and an increasing awareness ofthe cost of incentive programs argue for a selective stance. Weexpect that market conditions could become more favourable in2009.

Michael McNamara, Equity Analyst44 207 029 8680, [email protected]

Paul Clegg, CFA(212) 284-2115, [email protected]

Laurence Alexander, CFA(212) 284-2553, [email protected]

Please see important disclosure information on pages 208 - 210 of this report.

Michael McNamara, Equity Analyst, [email protected], 44 207 029 8680

Key Themes in Clean Technology

Clean Technology encompasses a wide range of industries and business models that stand to benefit from powerful secular trends in favor of more efficient use of resources in the face of rapid demand growth in the emerging economies that is stressing energy and water supplies. The main areas we focus on are alternative energy (wind, solar, biofuels, geothermal, etc.), carbon sequestration, green chemistry, and water. While Clean Technology is often viewed as a euphemism for alternative energy (aka oil or electricity substitutes), we approach the sector more broadly, looking for companies that that address constraints in feedstocks, energy inputs and water supplies, among others. In a world increasingly sensitive to carbon emissions, in some contexts efficiency can capture more value than new production. Make haste, not waste.

This kind of more holistic approach to the sector helps identify certain themes where the exceptions might be more interesting than the rule:

• It’s not the Internet. The Internet was characterized by bouts of significant capital investment followed by rapidly improving network economies (where each new user, at low incremental cost, added more value to the whole). The Clean Technology sector, in contrast, is dominated by companies that succeed or fail on a project basis. There are a few intriguing exceptions (smart metering, for example, or biotech traits).

• Asset plays vs. new technologies. It is critical, in our view, to distinguish between companies that are focused on building productive assets, and companies that have “asset light” strategies supported by proprietary technologies and multiple partnerships. The former can be expected to burn cash until the next projected plant is not supported by reinvestment economics; the latter can be expected to generate high free cash flows, with much of the project risk carried by other parties, albeit often with higher risk and or greater uncertainty of success. Ethanol plants compared to enzyme suppliers would be an example.

• It’s all about the regulations. As a general rule, regulatory initiatives should be the main driver for relative performance in the sector, much as it is for many other sectors of the economy (banks, HMOs, insurance companies, etc.). Credit market risk, feedstock prices, end-customer adoption rates, disruptive technologies—the most frequently cited risk factors can all be settled by the right incentive program. Breakthrough inventions, new process yields and efficiency levels, a judicious capital structure, favorable arbitrage economics, even significant partnerships—in short, favorable catalysts for shares can be thwarted by uncertainty over the direction of regulations. This is particularly true for asset plays, where subsidies can support projects for extended periods. Keep in mind, however, that volatile regulations, even if favorable, are likely to suppress valuation multiples due to the perception of uncertainty.

• It’s all about the energy costs. For many analysts, most of the action is focused on the price of power to end users, and whether companies are providing viable alternatives to electricity base load or maximizing electricity generation or transportation fuel (diesel or gasoline). We believe, however, that significant operating leverage can be found off the grid as well.

• Opportunities created by feedstock, production, or technology bottlenecks. In our view, this can be as potent a driver for share performance as regulatory fiat. Of course, there is a feedback loop involved: regulatory decisions contributed to the commodity landscape, and volatile commodity prices eventually prompt a policy response. For example, inexpensive corn encouraged biofuel subsidies, and now biofuel subsidies are under threat due to elevated corn prices. Similar dynamics have played out in a wide range of materials industries (e.g., silicon in 2006–08 and bearings in the wind industry). The sweet spot, in our view, occurs when regulatory initiatives are stymied by a technical or supply bottleneck which only one or a handful of companies are capable of solving.

• Funding has its cost. While it is easy to be seduced by new technologies, the funding arrangements are highly uncertain due to the capital intensity or the ongoing financial commitments required for the feed-in tariff subsidy model. We stress that investors should always review both the immediate capital cost of deployment of a new technology as well as the long-term liabilities that can be inherent in deployment.

These observations lead to a few rules of thumb for stock selection within the sector:

• Favorable regulatory trends: While the debate over climate change appears largely settled, we believe the debate over regulatory initiatives (carbon trading and taxes, sequestration, and favored technologies) is only beginning. Low carbon or energy-efficient technologies could see an accelerated adoption cycle in the appropriate regulatory environment.

• Sustainable: Companies that can reach profitability without regulatory support should do well. Business models need to be resilient, as “unforeseen” events are all too frequent. Examples of business model risks that have impacted shareholder returns in recent years include the diversion of LNG shipments to Europe and Asia hurting returns on U.S. LNG projects, the distribution bottlenecks that constrained the U.S. corn ethanol industry, or the difficulty of earning high returns on capital supplying components to commodity industries.

Clean Technology



• Scalable: New technologies need to be cost competitive and scalable, so as to earn an adequate return on the initial investment in R&D and process know-how.

• Credible path to profitability: Some clean technology areas involve attempts to earn high returns on capital supplying components to commodity or near commodity industries: we believe these are more vulnerable to competition than business models with a more robust approach to capturing value. We view pricing power, either from better technology or ongoing product bottlenecks, as a more sustainable driver of premium valuations than volume growth. We also prefer companies that seek to innovate to avoid commoditization of their current product offering.

• Higher growth, higher return. In our view, capital intensive projects could need to be “de-risked” in order to obtain project finance. This could constrain the returns available to equity holders, particularly for projects that combine new technologies in ways that have not been attempted before. Technology suppliers into commodity businesses could also see pressure on returns if they fail to innovate.

With this in mind, we recommend investors focus on the following criteria when evaluating opportunities in the highest-profile clean technology niches:

• Solar: Access to high-quality feedstock materials, cost reduction path to grid parity, the ability to effect cost reductions across the value chain and ability to secure financing.

• Wind: Proven technology providers with a roadmap to next generation solutions, reliable deliveries (turbine manufacturing), ability to secure project planning permission and financing (developers).

• Biofuels: Feedstock supply; distribution logistics; conversion efficiency; attractive ratio of $/BTU in vs. $/BTU out; ability to arbitrage favorably against higher-cost and alternatives.

• Industrial biotech: Scarcity value of technology; value proposition vs. existing petrochemicals without subsidies; geographic diversity; equal or better in terms of performance than rival materials.

• Water: Scalable technology; pricing power in an environment of rising water prices; volume growth driven by secular trends towards more efficient use of both water and energy.

Clean Technology



How to Play It

REC REC NO Solar

Rated Buy; Unique combination of strength in the preferred upstream silicon and wafer segments combined with potential cost reduction from FBR deployment, insulated from deteriorating module ASP.

Vestas VWS DC Wind

Rated Buy; World’s largest supplier of turbines to the wind industry where growth and pricing forecast considered less at risk, also the market leader in large scale turbines.

SunPower SPWR Solar

Rated Buy; Manufacturer of highest efficiency PV modules commercially available. A rare vertically integrated play that we believe will be able to grow profitable across cycles.

Shares to Consider

Energy Conversion Devices

ENER Solar

Rated Buy; One of only large scale manufacturers of thin, flexible modules well-suited for BIPV. Faces little near-term competition selling into high BIPV tariff markets over next few years.

Westport Innovations WPRT Alternative

Fuels

Rated Buy; Westport's CNG/LNG engine technology for heavy duty trucks provides a pure play on the adoption of natural gas engines to generate economic savings while reducing greenhouse gas and particulate emissions.

Solon SOO1 GR Solar

Rated Underperform; Potential margin squeeze from mismatch between solar cell costs and declining module ASP while working capital demands remain a concern.

Ascent Solar Technologies ASTI US Solar

Rated Hold; ASTI in the early stage of development and requires additional capital to fund future growth while significant execution risk remains. ASTI’s path to profitability may be unclear in an uncertain demand environment.

Nova BioSource NBF Biodiesel Rated Hold; Potential credit squeeze due to working

capital requirements and a tight lending environment.

Shares to Avoid

Where does the money come from?

From a top-down perspective, we believe investors should be sensitive to the fact that policymakers need to make some tough choices as to priorities over the next couple of decades.

At a recent conference we heard one presenter describe the projected investment in biofuels as “larger than the Manhattan and Apollo projects combined”—as if this is a good thing. To help put this in perspective, we believe it is worth juxtaposing some of the many investments we frequently hear the United States “must do” against the Congressional Budget Office’s (CBO) base-case forecast for the federal budget.

It is important to define the source of the funding. Some funding is a direct transfer from the federal general budget to the target industry while other solutions place the bill on the general population. Alternatively, the cost is not a direct cost but an opportunity cost where potential tax revenues are reduced via tax credits. Our views are relatively agnostic as we recognize the impracticability of applying a single formula across a variety of different tax and energy price regimes.

To the extent clean technology investments by governments might have to compete with other priorities, one way to frame the issue is in terms of the “fiscal gap,” or the relative scale of each funding initiative as a percentage of GDP. The following table compares various projections for investments that the United States should commit to —

Clean Technology


Company Symbol Industry Comment


whether upgrading basic transportation infrastructure, investing in clean technologies, fixing the Alternative Minimum Tax, or supporting the U.S. mortgage market under the Troubled Assets Rescue Plan (TARP) proposal. To some extent, certainly, these estimates lead to apples-to-oranges comparisons, due to their differing time horizons. The estimate of normalized impact to GDP assumes the investment is made over a 75-year time period, whereas certain infrastructure estimates are based on only a 20-year horizon. As such, if anything our estimate for the total cumulative impact to GDP from the proposed initiatives in the table below is likely too low. Even so, the prospect of initiatives that could expand the U.S. fiscal gap 125%–175% suggests that either 1) policymakers would need to make some hard choices to establish priorities, 2) innovation needs to dramatically reduce the cost of some of these initiatives, or 3) standards of living could come under pressure over the next 20–30 years. Politically, the second option would likely be the most appealing, in our view, and therefore likely the most subsidized.

EXHIBIT 1: FISCAL GAP (% OF FUTURE GDP) IMPLIED BY VARIOUS LONG-TERM COMMITMENTS

Commitment $ trillions % of future U.S. GDP Source Fiscal gap based on current budget scenarios 12.2 1.70% CBO including

Medicare Part D 6.5 0.90% CBO Social Security Deficit, 75 years 4.3 0.60% CBO

Other proposed commitments Clean Technology

Investments to abate climate change 6.0 0.84% Alliance Bernstein Alternately

U.S. GHG emissions abatement by 30% 1.1 0.15% McKinsey Build-out of 20bn gal/year of cellulosic ethanol 0.1 0.02% Jefferies, $6/gal capex EPA oversight of carbon cap & trade, first 10 years 0.1 0.02% CBO Total 1.3-6.0 0.19%-0.84%

InfrastructureU.S. Infrastructure, ex-bridges, highways, transit, rail & water, extrapolated from 5-year need (2005 estimate) 1.9 0.27% CBO, AASHTO, ASCE U.S. highways, transit & rail infrastructure gap, 20 years, extrapolated from 30-year need (2007 estimate) 3.0-4.1 0.42%-.57% SAFETEA-LU* U.S. drinking water & wastewater investment gap, 20 years 0.5-0.6 .07%-.08% EPA, ASCE Other U.S. water infrastructure (industry estimates) 0.5-0.7 .07%-.10% ITT

International water infrastructure 1.8-2.0 .25%-.28% ITT, World Water

Council Total 7.7-9.3 1.08%-1.30%

Other Policy ChoicesIndexing AMT to inflation 5.0 0.70% CBO Extending expiring income tax provisions 5.0 0.70% CBO Investments in oil production (globally, per IEA) 5.4 0.75% IEA Stay In Iraq, 75 years 3.0 0.42% CBO TARP: Proposed $700bn bailout for U.S. housing market 0.7 0.10% Media reports Total 19.1 2.67%

Total proposed commitments, above current unfunded mandates 28.1-34.4 3.94%-4.82%

Source: Jefferies & Company, Inc. estimates, ITT, World Water Council, CBO, SSA, AASHTO, ASCE, SAFETEA-LU (* extrapolated from 30-year forecast)

None of this is designed to suggest that funding will not ultimately be made available for Clean Technologies but rather to stress the hard choices policymakers face. These hard choices can lead to delays in implementing incentive programs as well as raising the potential that scarce resources could be funneled into more politically appealing choices potentially at the expense of both alternative programs as well as the environment.

Additional factors

At the state level, funding initiatives in the near term could be complicated by declines in revenue from property taxes and sales taxes. Moreover, state initiatives can be inconsistent when federal matching funds are unavailable, particularly for investments that are less tangible. Water infrastructure investments, for example, tend to be a lower priority than highly visible highway and bridge repairs.

Clean Technology



A related issue is whether state or federal regulators intend to set a price of carbon to achieve specific policy goals. In the near term, the U.S. appears to be trending towards a cap-and-trade system. Market-based pricing, however, may need to be supplemented by additional funding in order to establish sufficient incentives for changes in consumer behavior. We estimate, for example, that a $25/t CO2 price would only increase U.S. power prices by 14%, on a national average, although other significant factors, such as grid reinvestment and rising fossil fuel prices, could also have an impact.

EXHIBIT 2: CO2 IMPACT ON POWER PRICES

Coal Natural Gas Feedstock lbs CO2/MMBtu 205 135Reference coal Btu content/lb 12,000 -lbs per MT 2,205 2,205 MMBtu/MT 26.5 -MT CO2/MT Coal 2.5 -

Fuel cost per short ton/MCF $100.00 $7.50 Cost per MT CO2 $25.00 $25.00 Heat rate of plant (Btu/Kwh) 10,000 7,000 per KWh Coal/NG cost $0.038 $0.053 per KWh impact of CO2 cost $0.023 $0.011

Avg national price per Kwh $0.10 $0.10 Approx % generation coal 50% 50%Approx % generation natural gas 20% 20%Weighted average impact on power prices $0.014 $0.014 % increase in prices 13.8% 13.8%

Source: Jefferies & Company, Inc. estimates

One last consideration complicating the outlook is that government funding decisions could be driven by theoretical considerations, particularly arguments involving the energy return on alternative energy projects (EROEI). In its simplest terms, EROEI attempts to analyze whether the energy generated is worth the amount of energy one puts into the process. In practice, it is a metric that has an impact on investor sentiment, but to the extent that investors rely on EROEI models for an investment case, they make themselves more vulnerable to volatility induced by long-term assumptions that can be difficult to verify or benchmark in a consistent fashion.

The basic rule of thumb when considering EROEIs is that, when they decline, it takes more energy to maintain the same level of economic activity. For example, moving from a fuel with an EROEI of 20:1, where some analyses peg Middle Eastern oil, to 5 or 6:1, where some models place Alberta’s tar sands, implies a 16% increase in the energy required to maintain the same level of economic activity. This may be fine from an energy producer’s perspective (who wouldn’t want to capture a growing share of GDP?), but would likely prove unsustainable politically so long as more efficient alternatives exist, as governments want to minimize infrastructure investments. At least that’s the theory.

EROEI models, however, run into three fundamental critiques. First, they tend to ignore the value created by shifting from an energy source such as uranium ore to a different form such as transportation fuel or a cell phone battery. Second, there is a boundary problem: do you judge biofuels starting from when you plant the seed, or do you factor in the energy used to clear the land? Third, EROEIs often reflect a blend of steps supported by (less efficient) renewables and steps supported by hydrocarbons. This can make a value chain look more energy efficient than it would be on a consistent basis. Finally, high EROEI processes may face other constraints, such as the supply of rare metals, water, or other process inputs.

Clean Technology



EventJefferies' U.S. and European Clean Technology research teamshave collaborated to provide an overview of the current state andbackground of the solar industry as well as the various risks andopportunities investors face in the sector.

Key Points• Incentives Dominate. Uncertainty surrounding several macro

factors remains a recipe for continued volatility in solar shares.We expect incentives to continue to play a roll in stimulating PVadoption well beyond the threshold of "grid parity." Currently,we are focused on developments in the Italian and, to a lesserextend, the Greek markets as well as the recent passage of anexpanded and extended U.S. investment tax credit (ITC).

• Modules Galore? New polysilicon manufacturing capacityappears poised to help create abundant module supply andraises questions about the market's ability to absorb newmodules without price declines that are faster than incentivedigressions in key markets. While precise inflection points forsilicon or module availability and price changes are difficult topredict, both our U.S. and European analysts build in rapidprice decline assumptions in 2009.

• Cost Reductions. We think the PV sector is rife with costreduction opportunities, which could allow margin preservationin a declining module price scenario, depending on the paceand level of newly introduced incentives. Two key drivers ofcost reductions among silicon-based cell and modulemanufacturers are reductions in silicon costs and improvementsin conversion efficiency levels (the percentage of the Sun'selectricity generation potential harnessed). In both cases thesituation is encouraging for many players.

• Competing Technologies Gaining Ground. We expecttraditional crystalline silicon PV to maintain its dominant positionin solar markets for some time. Yet, newer technologyapproaches such as thin film, solar thermal (CSP) andconcentrating solar (CPV) are attracting considerable attentionand capital and many business plans are calling for significantincreases in production from pilot stages. A few thin filmproducers (e.g., Unisolar and First Solar) are already in largescale commercial manufacturing with impressive cost results,while others are experiencing "growing pains" related to scalingcommercial-scale production operations.

October 10, 2008

Clean TechnologyEnergy Generation - Solar

Clean TechnologySolar Primer

Investment SummaryWe see macro factors such as incentives and the availability ofraw material feedstocks as the most significant drivers for solarshares' performance. We expect rapid volume growth inpolysilicon and modules to drive down solar system prices fasterthan subsidy digressions in key markets. Well-funded businessmodels with strong cost-reduction levers may benefit.



James Harris, Equity Analyst44 207 029 8691, [email protected]

David Paek(212) 284-2175, [email protected]


Solar – A Secular Growth Story, but Invest Selectively

In the context of growing concern about climate change, rising energy costs and awareness of the potentially negative geopolitical externalities of current energy policies, we believe the political environment is ripe for further policy action to incentivize investments in renewables such as solar PV (photovoltaic), as a part of a portfolio of “clean tech” energy solutions. We see the implementation of much larger pools of PV incentives as highly likely (even if incentive levels in key markets step down year-over-year), driving rapid growth of PV adoption. We estimate that PV installation volumes could grow at a CAGR in excess of 50% through the end of the decade.

Despite this rapid growth potential, PV margins and returns on capital remain concerns, given the potential for mismatches between capacity expansion plans for solar, which can require long lead times, and incentive funding, which can be intermittent and subject to political compromise. Without incentives in key markets, most investments in PV installations would not be made, in our view, as PV is not yet cost-competitive with grid-metered electricity prices in most markets or with traditional energy generation sources. Thus, political fiat on specific regulations or incentives is among the most significant determinants of investor returns in solar, as are the availability of raw materials and other resources, including financing. In many markets, the political unwillingness to establish incentive mechanisms and programs with long-term visibility creates considerable near-term uncertainty and share price volatility around what appears to be shaping up a strong long-term secular growth story.

In the context of this high volatility risk-opportunity set, we believe investors should focus on large, well-capitalized names with strong cost-reduction roadmaps and well-funded business models. In Europe, we suggest that investors selectively build exposure to companies with less direct exposure to module ASP deterioration and with visible cost reduction strategies. We believe that REC (REC NO, Price Target NOK213, Buy) fits this description well with its commanding position in the silicon and wafer segments as well as the pending roll-out of its low cost FBR silicon production technology. In the U.S., we believe SunPower (SPWR, $110 Price Target, Buy) is well-positioned to ride out various possible market scenarios while maintaining strong operating margins through it vertical integration strategy. The company appears well positioned to take advantage of a potentially large utility scale market in the US towards the end of the decade. We believe Energy Conversion Devices’ (ENER, $96 Price Target, Buy), ability to supply a still unique, differentiated product that qualifies for very attractive building integrated (BIPV) feed-in tariffs provides a degree of protection against the primary concerns of most solar investors in the near-term (e.g., wafer pricing) and medium-term (e.g., oversupply and reduced tariffs). Moreover, ENER appears likely to experience strong operational momentum and rapidly expanding gross margins in the coming quarters.

EXHIBIT 1: GLOBAL SOLAR SECTOR COVERAGE

Current PriceCompany Analyst Symbol Price Target Upside Rating 2007A 2008E 2009E 2008E 2009ESolarWorld McNamara SWV GR €18.73 €37 98% HOLD 18.9x 15.2x 11.2x 30.1x 22.2xSolon McNamara SOO1 GR €21.3 €35 64% UNDERPERFORM 7.4x 13.3x 7.1x 21.9x 11.7xErSol McNamara ES6 GR €102.0 €58 -43% HOLD 118.6x 29.2x 22.2x 16.6x 12.6xREC McNamara REC NO NOK78 NOK213 175% BUY 28.7x 19.2x 9.6x 52.7x 26.3xQ-Cells McNamara QCE GR €34.0 €69 103% HOLD 27.4x 21.5x 11.7x 43.7x 23.7xPV Crystalox McNamara PVCS LN p107.8 p252 134% BUY 12.1x 11.6x 6.0x 27.1x 14.0xAscent Solar Clegg ASTI $3.97 $14 253% HOLD N/M N/M N/M N/A N/AEvergreen Solar Clegg ESLR $3.26 $13 299% BUY N/M N/M 10.5x N/A 41.9xEnergy Conversion Devices Clegg ENER $34.06 $96 182% BUY N/M 121.6x 20.9x 342.9x 58.9xSuntech Clegg STP $22.18 $70 216% BUY 18.6x 13.4x 8.2x 42.4x 26.0xChina Sunergy Clegg CSUN $4.01 $15 274% BUY N/M 13.4x 4.9x 50.0x 18.3xSolarfun Clegg SOLF $5.95 $22 270% BUY 14.2x 6.0x 3.7x 22.0x 13.8xSunPower Clegg SPWR $60.75 $110 81% BUY 48.2x 28.0x 17.9x 50.7x 32.4xEmcore Lau EMKR $3.72 $15 303% BUY N/M N/M 9.3x N/A 37.5x

Prices as of Oct. 9, 2008 Weighted Average 32.7x 26.6x 11.0x 35.7x 20.8x

P/E at Price TargetCurrent P/E

(Jefferies Ests)

Source: Bloomberg, Jefferies’ Estimates

For your reference, we have also provided Exhibit 2, illustrating the market cap and valuations of selected solar sector companies. While not all solar companies are included, this list comprises the largest and most liquid solar investments currently available.

Energy Generation - Solar



EXHIBIT 2: SELECTED SOLAR VALUATIONS

Company Analyst SymbolCurrent

Price

Shares Outstanding

(MM)Market Cap

(MM)Price

Target Rating P/E 07AP/E 08E P/E 09E

12/31/06 to

CurrentErSol Michael McNamara ES6 (Ger) €102.00 10.7 €1,094 €56 HOLD 120x 26x 21x 124%PV Crystalox Michael McNamara PVCS (UK) p107.75 416.7 p44,902 p252 BUY 14x 9x 8x IPOQ-Cells Michael McNamara QCE (Ger) €34.00 82.2 €2,793 €69 HOLD 24x 18x 11x 0%REC Michael McNamara REC (Nor) NOK77.5 494.3 NOK38,309 NOK213 BUY 29x 21x 10x -32%SolarWorld Michael McNamara SWV (Ger) €18.73 111.7 €2,093 €37 HOLD 19x 13x 10x -21%Solon Michael McNamara SOO1 (Ger) €21.30 12.5 €267 €35 UNDERPERFORM 6x 7x 7x -9%Aleo Solar Not Covered AS1 (Ger) €5.75 13.0 €75 NA NC 8x 5x 5x -15%CentroSolar Not Covered C3O (Ger) €3.69 14.5 €54 NA NC 37x 6x 4x -59%Colexon Not Covered HRP (Ger) €3.80 5.1 €19 NA NC n/m 6x 5x IPOConergy Not Covered CGY (Ger) €3.28 35.1 €115 NA NC n/m n/m 64x -93%Phoenix Solar Not Covered PS4 (Ger) €27.02 6.7 €181 NA NC 11x 9x 9x 76%Renesola Not Covered SOLA (UK) p203.00 135.7 p27,540 NA NC 12x 6x 5x -52%SMA Solar Not Covered S92 (Ger) €40.80 34.7 €1,416 NA NC 4x 14x 11x IPOSolaria Not Covered SLR (Sp) €3.09 101.1 €313 NA NC 3x 4x 5x IPOWacker Chemie Not Covered WCH (GR) €69.45 52.2 €3,622 NA NC 8x 7x 6x -29%Ascent Solar Paul Clegg ASTI (US) $3.97 18.4 $73.17 €14 HOLD n/m n/m n/m 37%China Sunergy Paul Clegg CSUN (US) $4.01 39.6 $159 $15 HOLD n/m n/m 4x IPOEnergy Conv. Devices Paul Clegg ENER (US) $34.06 45.7 $1,555 $96 BUY 70x n/m 10x 0%Evergreen Solar Paul Clegg ESLR (US) $3.26 164.7 $537 $13 BUY n/m n/m 8x -57%SolarFun Power Paul Clegg SOLF (US) $5.95 53.8 $320 $22 BUY 14x 6x 4x IPOSunPower Paul Clegg SPWR (US) $41.76 40.5 $1,693 $110 BUY n/m 19x 12x 12%Suntech Paul Clegg STP (US) $22.18 153.9 $3,414 $70 BUY 22x 13x 8x IPOAkeena Solar Not Covered AKNS (US) $2.21 28.8 $29.38 NA NC n/m n/m n/m -15%Canadian Solar Not Covered CSIQ (US) $11.16 35.6 $398 NA NC n/a 4x 3x 6%First Solar Not Covered FSLR (US) $115.28 80.0 $9,228 NA NC 81x 31x 17x 286%GT Solar Not Covered SOLR (US) $5.66 142.4 $805.93 NA NC n/m 8x 5x IPOJA Solar Not Covered JASO (US) $6.11 167.9 $1,026 NA NC 48x 6x 4x IPOLDK Solar Not Covered LDK (US) $19.91 106.5 $2,120 NA NC 15x 7x 4x IPOMEMC Not Covered WFR (US) $21.46 225.9 $4,848 NA NC 6x 6x 4x -45%Trina Solar Not Covered TSL (US) $11.24 29.3 $329 NA NC 7x 3x 2x -41%Yingli Not Covered YGE (US) $4.84 126.9 $614.19 NA NC 12x 5x 3x IPOEmcore John Lau EMKR (US) $3.72 77.7 $289 $15 BUY n/m n/m 17x -33%

Prices as of Oct. 9, 2008 Weighted Average 26.7x 12.3x 10.0x

Source: Reuters, Thomson First Call, Bloomberg, Jefferies’ Estimates

What to Expect in Solar – Executive Summary

While overall, we expect solar names to book strong revenue growth rates in 2008, we believe the uncertainty surrounding several macro factors remains a recipe for continued volatility in the shares. Key macro drivers, in our view, include resolution of uncertainty around incentive programs in key markets, as well as greater visibility on the likelihood of a potential surfeit of solar modules, as additional polysilicon supplies come online.

Incentives Still Dominate

We expect incentives to continue to play a key role in stimulating PV adoption and as a key determiner of module prices. Currently we are focused on the Italian and Greek markets both of which offer attractive highly attractive FIT incentives although many questions remain surrounding the potential scale of these markets in the coming 12-24 months. We are relieved to see the extension of the EEG in Germany which, although accelerates the rate of FIT digression, avoided any significant one-off adjustments and remains uncapped. In the U.S., Congress has just passed an unprecedented 8-year extension of solar Investment Tax Credits (ITCs) in a bill that could help to “kick start” a potentially enormous PV market in the US that has thus far performed below its potential. Certain provisions in the bill significantly expand access to the credit by eliminating caps and other restrictions on residential usage and allowing utilities to access the credit. In our view, any favorable impact on 2009 market growth is likely to be insufficient to forestall rapid global ASP declines, but it may provide a strong source of demand for subsequent periods (2010 and 2011).

Modules Galore?

Very strong returns among incumbent polysilicon manufacturers have resulted in a proverbial avalanche of announced investments in additional manufacturing capacity from existing players and new entrants. Large capital (and thus financing) needs, access to low-cost, reliable sources of power, expertise with developing and ramping chemical processes and access to scarce equipment sets have all been roadblocks slowing the progression of new polysilicon plants. Yet, we believe that even if success is limited only to first- and second-tier expansion projects, it would require a rapid acceleration in the rate of PV adoption to absorb new supplies. While an expansion of PV




subsidies could accomplish this feat, to us the pace of new market development does not appear sufficient to avoid a drop in module prices that is faster than planned reductions in incentive levels. Given the lack of accurate aggregate data available for silicon production being sold to the PV industry as well as a similar lack of reliable data on PV installations, makes it difficult to measure price elasticity of demand in this relatively new and highly subsidy-driven market and determining precise inflection points for silicon or module availability and price changes is difficult. On this very point our European and U.S. solar analysts agree that module ASPs could fall 15-20% in 2009 from current levels.

Cost Reductions

Given our expectation for rapidly declining PV modules prices, a focus on reducing costs is essential to differentiating business models among PV players. We think the PV sector is rife with cost-reduction opportunities, which could allow margin preservation, depending on the pace and level of newly introduced incentives. Two key drivers of cost reductions among silicon-based cell and module manufacturers are reductions in silicon costs and improvements in conversion efficiency levels (the percentage of the Sun’s electricity generation potential harnessed). In both cases the situation is encouraging for many players. Several China-based manufacturers are currently seeing their relatively low cost structures masked by very high polysilicon purchase prices that we expect to abate due both to a trend towards signing long-term contracts for silicon as well as our expectations for growing silicon supplies. Many European companies have already secured access to high quality silicon at attractive prices and are aggressively reducing their conversion costs. Other cost reduction opportunities include vertical integration of the value chain (particularly downstream into integration and installation), as well as volume buying and establishing dedicated sources of raw materials and equipment and bringing to bear industry best practices for cell designs, manufacturing processes and materials usage.

Competing Technologies Gaining Ground

While we expect traditional crystalline silicon PV to maintain its dominate position in solar markets for some time, newer technology approaches such as thin film, solar thermal (CSP) and concentrating solar (CPV) are attracting considerable attention and capital and many business plans are calling for significant increases in production from pilot stages. A few (e.g., Unisolar and First Solar) are already in large scale commercial manufacturing with impressive results (e.g., low costs). Yet, many of these technologies have experienced or are experiencing “growing pains” related to scaling technically workable pilot projects to commercial scale production operations. We are also beginning to see a significant ramp in the number of turn-key thin film solutions with Applied Materials and Oerlikon Solar as leading suppliers of production equipment.

In the case of concentrating solar (both CSP and CPV), we believe there is considerable potential to create a lower-cost form of solar generation in certain regions at utility scale. Yet, we note that the concentrated capital investment requirements and early stage of development of these projects remain potential speed bumps. Moreover, these technologies may compete with tradition forms of low cost generation (coal, nuclear) and may require transmission upgrades, while PV can be more effectively used as a distributed form of generation for which the relevant economic comparison is metered electricity (i.e., a fairer comparison). Although we note that many PV modules currently being installed are part of solar power plants.

Access to Feedstock

Between 2004-2007, access to feedstock was a key determiner of growth as many solar companies were caught short by the German- initiated boom in solar demand and were left to scramble for polysilicon. Given the rapid expansion in silicon production to date and projected over the next two years, we believe the dynamic is likely to change. We do not see silicon shortages as a systematic limitation to industry growth which lead investors to simply buy shares of those companies with silicon. Rather, we suggest that investors should focus on companies that have secured access to high quality and reliable silicon supplies sourced from companies with a competitive cost base. In particular, we stress that many new entrants to silicon production face production costs 2X higher than traditional producers. These new entrants, and the downstream companies that rely on this feedstock, could come under pressure if prices begin to fall and high cost silicon/wafer/cell/module producers face eroding margins.

Picking the Winners…and Avoiding the Losers

In general, we believe investors should be cautious with the solar sector as we suspect that City/Street module ASP forecasts could deteriorate leading to weakening consensus estimates which could create a headwind for solar shares in a market already leery of high beta names. However, we do believe that investors should take




advantage of periodic weaknesses to accumulate positions in certain leading solar names, particularly those with significant cost reduction roadmaps already in place.

Company Symbol Comment

REC REC NO

Unique combination of strength in the preferred upstream silicon and wafer segments combined with potential cost reduction from FBR deployment, insulated from deteriorating module ASP.

PV Crystalox PVCS

Leading European wafer producer has secure, low cost silicon supply from Wacker and Tokuyama. Will launch in-house silicon production in cooperation from long time TCS producer Evonik (formerly Degussa).

SunPower SPWR US

Manufacturer of highest efficiency PV modules commercially available. A rare vertically integrated play that we believe will be able to grow profitable across cycles.

Shares to Consider


ENER US

One of only large scale manufacturers of thin, flexible modules well-suited for BIPV. Faces little near-term competition selling into high BIPV tariff markets over next few years.

Solon SOO1 GR

Potential margin squeeze from mismatch between solar cell costs and declining module ASP while working capital demands remain a concern

Ascent Solar ASTI US

ASTI in the early stage of development and requires additional capital to fund future growth while significant execution risk remains. ASTI’s path to profitability may be unclear in an uncertain demand environment.

Shares to Avoid

2008 YTD Review

The year 2007 marked a strong performance for solar overall as strong incentives in Spain combined with continued growth in Germany and other key markets provided momentum for the investor enthusiasm for the space. Yet, in early 2008 the market outlook quickly turned bleak, as investors, concerned about the impact of the looming credit crisis, shed high beta, high value names that require external capital to meet growth projections. The effect was compounded in our view over resurfacing fears about rapidly growing plans for investment in silicon and PV cell and module capacity as well as growing uncertainty over the resolution of incentive questions in the United States and Spain.

The current year has been one of mixed fortunes for the solar industry. Module prices have held up very well in 2008 as Spanish demand has proven to be 2-3X larger than most insiders expected. This better than expected module price combined with ongoing cost reductions in the production chain have lead to improving margins for most solar wafer, cell, and module producers. However, despite this strong financial performance, most solar shares have suffered significant declines in their share prices as investors, already wary of high beta names in a weak market, have expressed concerns that module ASP declines have only been postponed. Please refer to the following Exhibits 3 and 4 which illustrate performance of most key solar names in the European and US markets.




EXHIBIT 3: US SOLAR SHARES 2008 YTD

-100%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

01 - Jan 01 - Mar 01 - May 01 - Jul 01 - Sep

Shar

epr

ice

retu

rn(lo

calc

urre

ncy)

Energy ConversionDevicesEvergreen Solar

SunTech

China Sunergy

SunPower

Emcore

Trina Solar

LDK Solar

JA Solar

Solarfun

MEMC

First Solar

ENER +100%

Source: Datastream

EXHIBIT 4: EUROPEAN SOLAR SHARES 2008 YTD

-100%

-80%

-60%

-40%

-20%

0%

20%

40%

60%

01 - Jan 01 - Mar 01 - May 01 - Jul 01 - Sep

Shar

epr

ice

retu

rn(lo

calc

urre

ncy)

REC

Q-Cells

SolarWorld

PV Crystalox

Ersol

Solon

Renesola

Conergy

Solaria

Aleo Solar

Phoenix Solar

Centrosolar

Source: Datastream




EXHIBIT 5: YTD RETURNS IN THE SOLAR SECTOR

-100% -80% -60% -40% -20% 0% 20% 40% 60%

ErSol (acquired by Bosch)


PV Crystalox

SolarWorld

Renesola

First Solar

LDK Solar

Q-Cells

Solon

Sunpower

REC

SunTech

Colexon

JA Solar

China Sunergy

MEMC

Trina Solar

Evergreen Solar

Solarfun Power

Solaria

Conergy

Solar sector (market cap weighted)

NASDAQ

TecDax

Source: Datastream

EXHIBIT 6: SELECTED SOLAR SHARES PERFORMANCE VS WTI SPOT AND OIL SERVICES INDEX (OSX) (10/09/07 -10/09/08)

0

20

40

60

80

100

120

140

160

180

200

10/9/2007

10/23/2007

11/6/2007

11/20/2007

12/4/2007

12/18/2007

1/1/2008

1/15/2008

1/29/2008

2/12/2008

2/26/2008

3/11/2008

3/25/2008

4/8/2008

4/22/2008

5/6/2008

5/20/2008

6/3/2008

6/17/2008

7/1/2008

7/15/2008

7/29/2008

8/12/2008

8/26/2008

9/9/2008

9/23/2008

10/7/2008

JEFFERIES SOLAR INDEX WTI CRUDE OIL SERVICES SECTOR

Source: Datastream

Exhibit 6 illustrates the relationship between the oil price and solar shares. While there is not always a tight correlationbetween solar and oil prices, there is often a short-term relationship when oil exhibits sharp moves, especially whencombined with heightened sensitivity to pending or potential legislation that favors alternative energies. In short, webelieve investors reason that high gasoline prices at the pump are likely to spur legislators into action on renewablesgeneration.




In Exhibits 7 & 8 below we reflect the volatility in PE valuations given to various solar plays during 2007 and 2008 YTD. While estimate revisions no doubt played a role in some of the sharp movements in the space, we believe that investor sentiment around the space was a key driver. One interesting element in these charts is the relative stability of the European trading multiples, with the obvious exception of Conergy, when compared to their US-traded counterparts, which include several ADRs of Chinese PV manufacturers.

EXHIBIT 7: SELECTED SOLAR PE RATIOS IN THE US- A WILD RIDE

0

10

20

30

40

50

60

70

Jan - 07 Apr - 07 Jul - 07 Oct - 07 Jan - 08 Apr - 08 Jul - 08 Oct - 08

Evergreen Solar

Energy Conv. Devices

SunTech

China Sunergy

SunPower

Trina Solar

LDK Solar

JA Solar

SolarFun Power

First Solar

Average sector P/E(excl FSLR):

2007 : 18.0x2008 YTD : 16.2x

Source: Bloomberg

EXHIBIT 8: SELECTED SOLAR PE RATIOS IN EUROPE

0

10

20

30

40

50

60

70

Jan - 07 Apr - 07 Jul - 07 Oct - 07 Jan - 08 Apr - 08 Jul - 08 Oct - 08

REC

Q-Cells

SolarWorld

PV Crystalox

ErSol

Solon

Renesola

Conergy

Solaria

Phoenix Solar

Average sector P/E:

2007 : 14.4x2008 YTD : 14.0x

Source: Bloomberg

Solar M&A

Activity in the solar sector slowed in 2008 with relatively few successful and several cancelled/postponed IPOs. As Exhibit 9 shows, most M&A deals in the past two years were along verticals or bolt-on acquisitions in the equipment space. We do not expect to see rapid consolidation of solar cell and module manufacturers unless valuations continue to decline. Today the economics of building new production facilities are superior to those of purchasing an entire company that owns them. Moreover, tolling agreements are a lower-cost way of consolidating unused capacity in this market. We believe that backward and forward integration plays could continue to dominate with larger cell and module manufacturers seeking to lower costs of distribution and integration for the end user of their products and seeking to secure long-term access to their raw material supplies. Similarly, silicon




providers may continue to seek downstream opportunities to as a way to hedge against the potential shift in pricing leverage along the supply chain as more and more silicon projects are announced.

EXHIBIT 9: SELECTED SOLAR SECTOR M&A 2007 AND 2008 YTD

Company Target Description of target company CommentSuntech Nitol Manufactures key chemical components

(trichlorosilane gas) for the global solar industry from chemical chlorine and silicon gas facility in the Irtutsk region in Russia.

Suntech purchased a minority interest in Nitol for $100 MM USD. Before Suntech's investment in Nitol, it signed a six year deal for a committed supply of polysilicon.

Suntech Hoku Polysilicon producer. In June 2007, Suntech agreed to purchase $678 MM of polysilicon from Hoku over a ten year period. In Feb. 2008 Suntech purchased $20 MM of Hoku stock in a private placement deal raising $25 MM.

Suntech Shunda China based manufacturer of PV cellsthat is in the finishing stages of building apolysilicon plant in Jiangsu province thatwill have an initial capacity of 1,500metric tons.

In May 2008, Suntech acquired a minority stake in Shunda for $98.9 MM. Suntech previously signed a long-term silicon wafer supply agreement for 7GWs over 12 years.

SunEdison Renewable NRG

Oregon based solar company that supplies and installs solar panels.

SunEdison was attracted to Renewable NRG's regional experience in Oregon.

Good Energies Solarfun Manufacturer of solar cells and module products.

Good Energies a leading renewable energy investors increased its stake in SOLF from 6.3% to 34.7% in Dec. 2007.

Norsk Hydro Ascent Solar Manufacturer of thin-film Building Integrated Photovoltaics (BIPV)

Norsk Hydro, a large oil equipment and aluminum manufacturer, is also active in building systems business which will utilize BIPV solutions in its products offerings. Hydro increased it stake in ASTI to 35% in April 2008.

DC Chemical Evergreen Solar

Provider of solar modules, wafers, and cells.

DC Chemical received shares in Evergreen Solar so that Evergreen can receive committed long-term polysilicon supplies of 1GW over 6 yrs.

Quantum Asola German solar products manufacturer. Quantum, a company based out of Irvine CA that develops fuel cell based technologies, purchased a 25% stake in Asola in January 2007.

Applied Materials Baccini Italian based company that develops and manufactures metallization and test systems for the polysilicon manufacturing industry.

Applied Materials purchased Baccini for $330 MM in November 2007.

Applied Materials HCT Swiss based HCT Shaping Systems is a leading developer and manufacturer of equipment used in the manufacturing process of polysilicon wafers.

Applied Materials purchased HCT for $475 MM in June 2007.

Itochu Solar Depot California based solar company that develops and integrates solar-thermal systems and BIPV products.

Itochu, already a player in the downstream solar business in Asia, purchased Solar Depot to penetrate the US solar markets.

SunPower Solar Solutions Distributor of solar products and solutions based in Faenza, Italy.

SunPower purchased Solar Solutions primarily to penetrate the downstream solar business in Europe.

WorldWater & Solar ENTECH Provides advance concentrator solar technology for a variety of applications.

WorldWater & Solar and ENTECH completed their merger in January 2008.

First Reserve Gamesa Solar Based out of Spain, Gamesa Solar sells, develops, and constructs PV plants.

First reserve, a private equity firm investing in energy and energy related industries, completed its acquisition of Gamesa Solar in February 2008 for 269 Million Euros.

Robert Bosch GmbH

Ersol Integrated wafer & cell producer, silicon sourced from Hemlock and Wacker

First sign of a leading industrial company entering the solar sector via acquisition

Source: Jefferies International Ltd.




The only U.S. IPOs in solar in 2008 were Real Goods Solar and GT Solar, both of which has not performed well since pricing while in Europe SMA Technologies stands alone. IPOs launched in 2007 included a handful of Chinese solar cell and module manufacturers which listed in the U.S. and in London.

EXHIBIT 10: SELECTED SOLAR SECTOR IPOS 2007 AND 2008 YTD

Company Ticker MarketCapital Raised

(in MMs) DescriptionJA Solar Holdings JASO USA $225 JA Solar Holdings Co. Ltd. Manufactures solar

cells. The company sells its products principally to solar module manufacturers which integrate its products into modules and cells.

Jetion Holdings Ltd JHL LN UK £38 Jetion Manufactures solar cells and modules.

PV Crystalox Solar PLC PVCS LN UK £220 PV Crystalox Solar is a leading producer of silicon wafers for the solar industry and is planning to initiate silicon feedstock production.

Real Goods Solar RSOL USA $55 Real Goods Solar Inc. is a residential solar energy integrator. The company designs, install, and maintains solar energy systems.

Yingli YGE USA $319 Yingli Green Energy Holding Company Limited designs, manufactures and sells PV modules. The company also designs, assembles, sells and install PV systems that are connected to an electricity transmission grid or those that operate on a stand-alone basis.

SMA Solar Technologies S92 GR Germany € 362 SMA is one of the leading suppliers of high efficiency inverters to the solar industry

GT Solar USA $500 GT Solar is a leading producer of deposition reactors for Siemens based polysilicon production. The company also produces production equipment used in the manufacture of crystalline solar wafer and cells.

Source: Jefferies International

Sizing up the Solar Market

In 2007, we estimate that total production of solar modules was approximately 3 GW, of which the bulk were sold and installed by early 2008. This represents 60+% volume growth rate year-over-year. We estimate that current total installed worldwide MW of PV were approximately 8 GW at the end of 2007. This is less than the EIA’s estimate of planned natural gas generating capacity additions in the U.S. market for just one year (2008). (We note that natural gas peakers also have somewhat higher capacity factors than solar, making the comparison more stark).

EXHIBIT 11: SOLAR ONLY A SMALL PERCENTAGE OF TOTAL ENERGY PRODUCTION

2003 2010E 2015E 2020E 2025E 2030EWorld electricity consumption in MM Kwh 14,781,000 19,045,000 21,699,000 24,371,000 27,133,000 30,116,000Solar installed capacity (GW) 1.8 32.9 176.8 656.3 1,633.1 2,630.0Interim CAGR NA 51% 40% 30% 20% 10%Solar MM KWh, based on 1200 KWh/KW/Yr 2,194 39,438 212,108 787,544 1,959,660 3,156,053% solar generation 0.0% 0.2% 1.0% 3.2% 7.2% 10.5%

2003 2010E 2015E 2020E 2025E 2030EWorld electricity consumption in MM Kwh 14,781,000 19,045,000 21,699,000 24,371,000 27,133,000 30,116,000Solar installed capacity (GW) 1.8 32.9 176.8 656.3 1,633.1 2,630.0Interim CAGR NA 51% 40% 30% 20% 10%Solar MM KWh, based on 1200 KWh/KW/Yr 2,194 39,438 212,108 787,544 1,959,660 3,156,053% solar generation 0.0% 0.2% 1.0% 3.2% 7.2% 10.5%

Source: Energy Information Administration and Jefferies’ Estimates




EXHIBIT 12: PV INSTALLATIONS GROWTH

PV Installations

0

500

1000

1500

2000

2500

3000

2002 2003 2004 2005 2006 2007

MW

On Grid Off Grid

CAGR 46%

Source: www.solarbuzz.com & Jefferies estimates

Regional Segmentation

Most solar modules are produced in Europe, Asia and the United States, with China taking a growing piece of the pie mostly at the expense of Japan and to a lesser degree Europe. For solar end markets, Europe dominates while Japan has leveled off (even declined in 2007) and the U.S. shows the potential to become an industry giant, despite initial inertia. China holds promise, including for off-grid applications, but the incentive and/or regulatory structures necessary for rapid adoption may still take years to develop. Most silicon is currently produced in the United States and Europe, but a large number of new operations are coming online in China.

EXHIBIT 13: GLOBAL CUMULATIVE INSTALLED SOLAR CAPACITY BY REGION (2007)

Other EU5%

Rest of World3%

Germany49%Spain

8%

Japan24%

US11%

Source: IEA –PVPS Trends in Photovoltaic Applications 1992–2007 (www.iea-pvps.org)




EXHIBIT 14: INSTALLED CAPACITY GROWTH 2001–2005

Cumulative Installed PV

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

MW

EU N. America Japan Other Estimate

Source: www.iea-pvps.org & Jefferies’ estimates

EXHIBIT 15: GEOGRAPHIC DISTRIBUTION OF MANUFACTURING

Cell Production by Geography

0 200 400 600 800 1000 1200 1400

Africa & Middle East

Australia

India

Other Asia

USA

Other Europe

Taiwan

Germany

Japan

China

2006 2007

Source: Photon International & Jefferies’ Estimates




EXHIBIT 16: TOP 10 CELL PRODUCERS – THE JAPANESE LOSE GROUND TO CHINA AND THIN FILM

2007 2006Q-Cells 9.1% 10.0%Sharp 8.5% 17.1%Suntech 7.9% 6.3%Kyocera 4.8% 7.1%First Solar 4.7% 2.4%Motech 4.1% 4.0%SolarWorld 4.0% 3.5%Sanyo 3.9% 6.1%Yingli 3.4% 1.5%JA Solar 3.1% 1.2%Other 46.6% 40.8%

100.0% 100.0%

Source: Photon International

EXHIBIT 17: TOP 10 MODULE MANUFACTURERS

Top Module Manufacturers in 2007

0 50 100 150 200 250 300 350 400

Solon

BP Solar

Mitsubishi

SolarWorld

Yingli

Sanyo

First Solar

Kyocera

Sharp

Suntech

Top Module Manufacturers in 2007

0 50 100 150 200 250 300 350 400

Solon

BP Solar

Mitsubishi

SolarWorld

Yingli

Sanyo

First Solar

Kyocera

Sharp

Suntech


Growth Projections

Jefferies estimates that PV adoption could grow at an accelerated rate in the coming few years due to the increased availability of silicon with which to make cells and modules as well as the increasing success of thin film production processes, which have been gaining traction (please refer to Exhibit 18). We depict our growth forecast through 2009 as well as the announced growth plans for selected major cell providers in Exhibit 19 below. In 2008, we expect that the industry could grow at approximately 50% before perhaps doubling in the following year, due to a rush of new silicon and cell and module supply coming on line.

In our view, one simple forecasting problem relates to where to put all of the modules we expect to be produced. Based on historical and current adoption rates in key markets, one must assume very rapid acceleration of PV adoption in order to avoid building large amounts of inventory. As we do not see new incentives regimes expanding at a rate sufficient to absorb new module supplies, we remain convinced that prices will need to decline faster than incentive levels to stimulate sufficient demand through expanding IRRs. The precise timing of rapid price declines is difficult to target, given a lack of transparency around silicon production data points and the difficulty in estimating price elasticity of demand in a relatively new and highly subsidized market. Our European and US clean tech teams believe that module ASPs could fall 15-20% in 2009 from current levels.




EXHIBIT 18: PV CELL GROWTH FORECAST

Solar Cell Production (MW)

0

2,000

4,000

6,000

8,000

10,000

12,000

1999 2000 2001 2002 2003 2004 2005 2006 2007 2008E 2009E

Min

Max

Min

Max

Source: Photon International & Jefferies’ Estimates

EXHIBIT 19: SELECTED 2008 PRODUCTION PLANS

Selected Cell Production Plans (MW)

0 200 400 600 800 1,000 1,200

China Sunergy

Sanyo

Kyocera

SunPower

Yingli

JA Solar

Sharp

Q-Cells

Production 2008 Capacity End 2008 2007 Production





Solar Incentives & Other Drivers

Given that solar is not yet cost competitive versus traditional generation in most markets, government-sponsored incentives are required to promote solar investment. A well-designed initiative will provide both the security of longevity to attract investors to the arena and the incentives to reduce the up-front production cost of capacity. While forecasting the timing and effectiveness of incentives is tricky, we would suggest that investors focus on a few key markets, which should serve as catalysts to share price performance. Key issues to focus on will be the level of Feed-in-Tariff (“FIT”) digression in the renewal of the EEG (Erneuerbare-Energien-Gesetz) in Germany; creation of a post-September 2008 framework in Spain; and new state/federal incentive plans in the USA, including the extension of the 30% investment tax credit.

Below we describe the different approaches taken towards incentive structures in different countries. While Europe has largely opted for a FIT model, the United States has focused on various regional strategies that emphasize upfront one-time payments or tax credits to offset the initial costs of a PV system.

Broadly speaking, incentive schemes worldwide can be split into two camps, the feed-tariff model and the subsidy tax relief model. The reason for the difference in incentive programs is largely related to the tax systems in the individual countries.

Feed-In Tariff Model (Europe, S. Korea)

The feed-in model is best personified by the EEG introduced in Germany in 2004. Rather than focus on subsidizing the cost of the installation and/or reducing the electricity bill, it guarantees a high feed-in tariff to create an attractive financial investment. Under the feed-in tariff model, 100% of production is exported to the grid (to the local utility) at a tariff guaranteed for 20 years (in the German case), while electricity consumed is imported from the grid and normal electricity rates.

The cost of the higher rates paid to the owners of the PV system are borne indirectly by residential utility ratepayers, while industry and business are exempt. Thus, as overall FIT payments rise, overall retail electricity rates will also rise, although the use of solar generation (in effect, by the utility) will also allow for some offset to the utility’s cost of complying with carbon reduction targets. However, given that solar generation penetration rates remain low compared with total electricity generation, the impact is thus far nominal.

Many FIT regimes control the amount of money allocated to FIT through a cap on the total number of MW that can be cumulatively installed to receive that tariff. Once these “breakpoints” on total installations are reached, the tariff either steps down to a lower level (sometimes the average retail tariff), or must be replaced by another regime. In Germany, the FIT regime provides longer-term visibility by reducing the tariff rates by a pre-determined percentage on January 1st of each year, although this percentage is reviewed and altered periodically. In Spain, the second-largest PV market in 2007, an existing cap structure FIT scheme is on the cusp of being updated by a new version, although the details are unclear.

In the FIT model, the ability to access low-cost, upfront financing for a large portion of the module and installation costs is important. In the exhibits below, we outline feed-in tariff structures in key markets.

EXHIBIT 20: SELECTED EUROPEAN SOLAR FEED-IN TARIFF PLANS

Country Tariffs Digression Duration Tax Credit Other

Germany €0.44-0.47 / kWh 8-11% annually through 2011 20 Years No Not capped, but digression impacted by

market growth

Spain €0.32-4 / KWh Inflation – 1% 25 years No Capped at 500MW in 2009 & 400MW in 2010 – still pending

France €0.30 / kWh, €0.50 / kWh for BIPV N/A 20 Years 50%

(max €8000) Rooftop FIT, 500MW cap

Italy €0.35-0.40 / kWh 2% 20 Years No 1200 MW cap

Greece €0.40-0.45 / kWh 1% monthly 20 Years No

Developers could qualify for 40% capital subsidy from EU infrastructure funds. Cap is currently 500-700MW with potential for additional 750MW rooftop only cap

Source: Photon International, Jefferies International Ltd.




Subsidy/Tax Relief Model (United States)

The subsidy/tax relief model is best demonstrated by the incentive programs in the United States. In contrast to the FIT model, the tax relief/subsidy model provides upfront relief on the initial cost of a solar system, through a direct payment to the owner of the modules (usually at a state or local level), or through a tax credit (offset of owner’s income tax liability). In the U.S., several state and local authorities also provide “buy-down” payments equating to a $/watt figure. The most notable is California’s Solar Initiative, which allocated more than $3 billion to incentive payments in the form of either buy-downs paid on a per watt installed basis or a Performance Based Incentive (PBI) that pays out a subsidy per KWh generated for the first five years of service. Many other state-level incentive programs in the U.S. have experienced lumpy funding, which has limited their effectiveness as a consistent PV adoption tool.

Incentive programs in the United States are based on a process called net metering. This process requires the utility companies to buy electricity generated from solar systems at the retail price of electricity. This electricity is then fed back into the grid. In this arrangement, the customer that generates electricity would still buy his/her electricity from the grid; however, he would only pay for the difference between the amount of electricity produced and the amount of electricity consumed. In contrast to Europe, U.S. installers only have an incentive to match their solar installations production with their actual use of electricity. If a solar system generates more power than is used by the customer, then the utility will still buy the excess power, but instead of paying the retail price of electricity to the solar generator, the utility is only required to pay its cost to produce the electricity. The cost to produce electricity for the utility may be only one-third of the retail price. Therefore, in most cases, we do not think it makes sense to build residential or small commercial solar systems larger than the customer’s demand for electricity.

With tax subsidy systems, the ability to absorb the tax benefit of the subsidy is important and the industry tends to use equity partnership structures combined with purchase power agreements (PPAs) to allow entities with no tax appetite (or no desire to own the modules) to “monetize” the value of these credits. A PPA is simply a contract to purchase power from the producer over a period of time. When the purchaser of the power is a quality credit counterparty, this facilitates using high levels of project debt, which improves equity returns.

Other inducements like accelerated depreciation for tax purposes also play into the financial returns and effectiveness of the U.S. model, as do the value of Renewable Energy Certificates (RECs) related to meeting renewable generation targets under state mandates. In certain states, including California, New Jersey, and New York, cumulative incentives could subsidize two-thirds of the total costs of a new solar system. A breakdown of the incentives provided by the United States is discussed below.

Regional Incentive Developments

Over the coming years, we expect continued growth in PV driven by new incentive programs, the expansion of existing programs and the reduction of module and system prices, which we believe will boost returns to investors and stimulate additional demand. Key markets to watch for changes to incentives in 2008, include Spain, the U.S. and Germany. Japan is currently considered a self-sustaining market without national subsidies today, but has experienced only modest growth since the end of the last subsidy regime, which ended at the end of 2005 (in 2007 installations actually declined).

Germany – Current Market Leader

In 2005 the German PV market surpassed Japan as the largest solar market in the world based on MW of PV installed that year. The driver of this growth has been favorable Feed-In-Tariffs enacted by legislation commonly called the “EEG.” Current rates of return are low on a total project basis, but levered equity returns can still be quite attractive due to low interest rates. The current FIT is approximately €0.44–0.47/KWh, with the higher tariff being for rooftop systems and the lower for larger field installations.

The EEG has been renewed and the result was roughly in line with industry expectations. The FIT digression rate for small scale rooftop will be 8% while ground mounted systems and large scale rooftop will have a FIT digression of 10% in 2009 & 2010. All installations will have a uniform 9% digression in 2011. Additionally, there is a “growth corridor” which can lead to a -100 / +100 bp change to the digression should annual installations fall below or above the predetermined levels. While the digression rates may have been a bit higher than the industry would have preferred, it is key to note that there is currently no cap on installations in Germany.

Japan – Growth Post Incentives

Japan presently is one of the most mature PV nations, mainly because it was the first to introduce large incentives for PV systems. It currently has the second largest installed base of PV (Germany No. 1) and was the third largest market for new installations in 2007 (behind Spain). The majority of installations in Japan are residential PV




rooftop systems. Japan introduced rebates in 1994 to offset the high cost of solar at that time and due to the long-term nature of those incentives it was able to bring down the cost of PV by 72% in 10 years. At this point, the Japanese government has eliminated most of the incentives as the PV market has now reached maturity in this country although growth has slowed considerably. Recently, the government has suggested that it may re-start its solar incentive programs. Policy details remain uncertain and we do not believe that policy support will trigger significant increase in demand in 2009. However, we stress that Japan is a price elastic market that will respond favorably to the increased availability of lower cost modules.

The United States – Potential Long-Term Driver of Solar Growth

The United States was the fourth largest solar market in the world in 2007 but it has not approached its potential contribution to the solar industry, in our view and could slip to fifth place in 2008. Despite healthy solar industry growth rates, the country only gets a negligible percentage of total electricity production from solar PV (less than 1%). The U.S. incentive system is tax credit based at the federal level, as described above, while at the state level cash “buy-downs” of the cost of a system are paid to installers or purchasers of the system to offset the initial cost.

State Incentives. In total, more than 40 U.S. states have some sort of incentive program and have mandated net metering. The largest PV state, California, through the California Public Utilities Commission (CPUC) enacted this as the California Solar Initiative (CSI). The CSI is a long-term incentive program that will provide $3.2 billion in rebates to both residential and commercial installations of solar systems. The goal is to install 3000 MW of solar modules by 2016, although the initial success of the program suggests that the State could reach these goals much earlier.

The initial “buy down” rebate was slated at $2.80 per watt, but declines as certain breakpoints in the number of MW approved under the system are reached and is currently at $1.55 per watt. A performance-based-incentive (PBI) is also available, which pays out a fixed per KWh subsidy (currently $0.22/KWh) for five years for commercial systems in lieu of a buy-down. In certain locations, municipalities also offer financial incentives or attractive financing to subsidize solar installations as well. The City of Berkeley announced an innovative plan in 2007 to offer low-cost financing to homeowners installing PV systems, using public bond authority (i.e., passing through a very low interest rate to the purchaser). The means of repayment is a long-term adjustment to the purchasers’ property taxes, making the incremental periodic cost of the system modest and obviating the need for a large upfront payment, which can be a significant barrier to PV sales.

As Exhibit 21 indicates, California and New Jersey represent the vast majority of all PV installations in the United States, although bureaucratic difficulties with the funding of the New Jersey program caused that market to stall in 2007. Many other U.S. states have or are considering incentive programs that could help drive incremental adoption. We note that most of these programs do not have the same long-term funding horizon as the CSI.

EXHIBIT 21: GRID-TIED PV INSTALLED IN 2007 BY STATE

California59%

New Jersey11%

Nevada10%

Colorado8%

Connecticut1%Arizona

1%Haw aii

2%New York

3%

Massachusetts1%

Oregon1% Other

3%

Source: Prometheus Institute

Federal Incentives. While state-by-state incentive programs continue to proliferate, the key platform of nationalincentives remains the 30% U.S. investment tax credit (ITC). Historically this credit expired at 1-2 year intervalsand was restricted by a $2,000 residential cap, an AMT exclusion (payers of Alternative Minimum Tax did not haveaccess) and a utility exclusion (utilities did not have access). At the time of this report, the US Congress had justpassed a bill (HR 1424) which provides for an eight-year extension of the credit through 2016 with amendments to




remove the $2,000 cap as well as AMT and utility exclusions. The President has already signed the bill which was a part of a larger “bailout” package for financial institutions. We believe that this multi-year extension and expansion of the credit will allow the US market to develop more long-term business models around PV and significantly enhance the potential of the US market to help absorb the large number of modules we expect to come on line in the coming years. At 250-300 MW in 2008, we do not think that even substantial year-over-year growth is likely to forestall rapid price declines in 2009. Yet, in 2010 and beyond, we believe that the combination of increasing utility involvement in PV, rising electricity prices, a growing REC market and abundant sunshine in the US is likely to awaken substantial market potential in the US market. As we expect that utility scale involvement will be a substantial driver beginning in 2010 in the US market, we do not expect this demand to support strong ASPs (utilities require low cost installations), but it could help bolster demand to avoid rapid global ASP declines due to oversupply.

In the United States, we see three types of legislation having the potential to raise investor expectations on the speed of solar adoption. First, legislation that extends or enhances existing tax incentives for solar and other renewables could speed solar adoption. Second, a new national renewable portfolio standard (RPS) is within the realm of political reality and could boost demand for all forms of renewables, in our view. Third, CO2 legislation (in addition to higher fossil fuel prices and the cost of grid reinvestment) could incrementally increase the cost of electricity generated from coal and natural gas, helping to incentivize the adoption of renewables.

In Exhibit 22 below, we have summarized the previsions of the recently passed HR 1424 dealing with energy tax credits for renewables and other forms of energy and generation.

EXHIBIT 22: HR 1424 MAJOR PROVISIONS FOR ENERGY TAX CREDITS

Status: Enacted Oct. 3, 2008Provisions:Solar -8 Year Extension of 30% investment tax credit through Jan. 2017 for both residential and commercial solar projects

-Allows AMT tax payers to offset tax credits against their alternative minimum tax liability-Removes the $2,000 cap on residential solar investment tax credit-Allows electric utilities to claim solar investment tax credits

Other -1 year extension for wind, refined coal technologies, and geothermal projects

CREBs -Allows the creation of $800 million Clean Renewable Energy Bonds (CREBs) to finance renewable energy generation projects

R&D Tax Credit -1 year extension of the R&D tax credit

(H.R.1424) Emergency Economic Stabilization Act of 2008

Source: US Congress, Bloomberg, SEIA. * AMT = Alternative Minimum Tax

U.S. State Goals Indicative of Future Incentive Programs

As Exhibit 23 indicates, 36 states have established aggressive goals for renewable energy production, with many of the states calling for a minimum of 10% of energy production from renewable sources. In order to meet these goals, we believe many states are considering or have already implemented incentive programs for the adoption of solar, wind, geothermal, etc. We note that RPS standards tend to favor wind energy as the least cost solution for compliance, but many also have specific carve-outs for solar, as depicted in Exhibit 24.




EXHIBIT 23: RENEWABLE PORTFOLIO STANDARDS IN THE UNITED STATES

Source: www.dsireusa.org

EXHIBIT 24: STATE RPS STANDARD WITH SOLAR “CARVE-OUTS” IN THE UNITED STATES

Source: www.dsireusa.org

Spain – Next Driver of Industry Growth

Spanish demand surged in 2007-8 driven by the high IRRs on offer due to the combination of high FIT and excellent insolation in the Iberian peninsula. What had been targeted as a 370-400MW solar program in the Ley Decreto 661/2007 instead mutated into a 1500-2000MW beast as developers exploited loose planning restrictions and the twelve month “grace period” to rush through projects and capture favorable rates. The unexpected size of the program has lead to unexpected costs as well. The original program had an expected price tag of €200-250m annually over the next 25 years. Now the bill will likely be more than €1b annually for 25 years.

The new draft royal decree has not yet been finalized at the time of writing although it seems very likely that the FIT will be scaled back to the €0.32-34 / kWh range and, more importantly, a cap on FIT qualified installations will be imposed with the most likely outcome of 400MW annually in 2009-2010. While the returns in Spain remain attractive, the strict cap will limit this market.

Italy – The Next Frontier? Italy offers a 20-year, multi-tiered FIT with current rates of €0.35–€0.40/ KWh. The program includes a 2% annual digression and a 1200 MW cap. While initial bureaucratic friction on system




approvals and payments slowed initial progress there, anecdotal evidence suggests significant activity on new installations in 2008 and being planned for 2009 (although the precise levels of activity does vary substantially between sources). We believe that high electricity rates and abundant sunshine combined with attractive FIT could eventually make Italy among the strongest PV markets in Europe.

Looking to 2009, Italy is one of the key markets to watch. The high return potential could trigger a “Spain-like” rush to develop projects and provide support for module prices. While installation growth is strong (albeit from a low base), bureaucratic delays, often linked with local planning permission, has thus far hindered mass deployments.

France. Along with the aggressive wind targets, the French government is offering attractive incentives to solar investors. A rooftop system may earn a feed-in tariff of €0.30/KWh, while the investor can receive a tax credit for 50% of the system cost up to a maximum of €8,000 per household. True BIPV (Building Integrated Photovoltaic) applications where the modules take on a practical function as part of the building (e.g., awning, window, etc.) qualify for a €0.55/KWh FIT.

Greece. Initial enthusiasm around strong FITs in Greece has been stymied by bureaucratic difficulties on new project approvals and payments. We have seen evidence that the Greek government is keen to open the solar market by lifting the effective blockade on permissioning while in return increasing the FIT digression to 1% monthly. At the time of writing, the pertinent legislation was still pending. Greece’s excellent solar resources and high electricity prices could make the market a strong addition to European demand.

China – Long-term Growth Driver? We believe that China holds long-term prospects for solar demand, given government and rising public concerns about the level of pollution being created by the country’s rapid industrialization. However, initial enthusiasm over a Renewable Energy Law enacted in 2006 waned after it failed to catalyze action on solar adoption on a broad basis. We believe that this national “law” was somewhat more of a proclamation and that the provincial and local authorities are the means by which real incentives to drive solar will be accomplished. Thus far, only a few relatively small local programs have been launched. We believe over time, the need for rural electrification in Western China could create significant demand for off-grid solar installations.

Other World. Numerous countries including Mexico, Canada and Czech Republic have announced incentive programs for solar. Korea shows particular promise, with strong FIT structures and high energy costs.

Non-Financial Barriers

The emergence of financial incentives is, of course, critical to the success of solar. However, it is important to review some of the non-financial barriers to mass deployment of solar systems. Delays and bureaucratic hassles involved in installing solar systems can act as a fairly significant impediment. Initial bureaucratic tendencies in new solar markets can stall progress initially around project approvals and FIT payments. Other potential roadblocks include:

Net Metering Rates. This refers to the price at which excess electricity is sold to the grid and is not relevant in markets that have a feed-in tariff-based incentive plan. Unattractive net metering rates (less than the retail tariff) can significantly reduce the financial incentive to install solar power, particularly in the residential market where the system may generate more power than the residential daytime load.

Net Metering Caps. Regional grids may have limits on the amount of distributed generation and/or net metering that can be installed in a certain region. This situation occurred in parts of California and forced a halt to installations until the cap could be eliminated.

Planning Permission. Many communities require planning permission in order to install solar panels on rooftops and can cause delays in installation.

Grid Connection. Some regional/national grids have detailed specifications for generators seeking to dispatch to the grid. These specifications are often designed for central generation plant (200 MW +) and could place a heavy burden on homeowners.

While these problems may seem daunting, they can all be overcome although the process will certainly take some time. Additionally, many solar installers are seeking to provide much of the planning and permitting required as part of the sales package both to increase their potential revenue as well as incentivize sales.




2008 Supply / Demand

Our growth model forecasts very high rates of adoption over the next few years (50% in 2008, potentially ~100% in 2009) driven by increased availability of modules, combined with the expansion of available subsidy pools for solar. Our analysis centers on silicon availability and concludes that increased investment in upstream capacity from new and existing players could drive module and system prices down faster than incentive level step-downs in key markets to create attractive returns for project investors and drive faster adoption.

Silicon Supply – A Gaiting Factor

The availability of polysilicon has been a gaiting factor on production of crystalline silicon modules. While the industry historically lived off the scraps of the electronics industry, solar currently makes up more than half of all silicon demand and despite more efficient use of the material solar silicon demand is growing substantially faster than electronics based demand, which has been growing at approximately 5%–8% p.a., we believe. In order to secure long-term supplies of silicon, most PV manufacturers have signed long-term contracts with polysilicon producers. Yet the long time horizons necessary to build and ramp a successful polysilicon production facility conflict with the near-term attractive economics of current PV incentives, particularly in Europe. This has caused wafer, cell and module manufacturers to scramble to source virtually all available polysilicon, driving current spot market prices of the material up to $500/kg for high-purity silicon, which then is often blended with lower-quality scrap material to make ingots of a sufficient quality to manufacture more cells and modules. We note that recent reports of declines in spot market pricing has been accompanied by explanations that the drops were less due to increased availability of the raw material, as much as the lack of remaining scrap material with which to blend. Long-term contract prices of silicon have also risen from levels of $30–$50/kg in 2004 to rates well higher than $100/kg for some contracts, although most long-term contracts we are aware of incorporate gradual price declines in silicon pricing to levels below $50/kg by 2012–2014. As silicon (ingots) can make up as much as 70%–80% of the cost of a module for some producers, significant cost leverage to declining prices appears likely.

EXHIBIT 25: POLY-SILICON WAFER PRODUCTION

Source: REC 2005 Annual Report

We believe that the top seven incumbent polysilicon producers made up more than 90% of polysilicon production in 2007. However, new entrants are attempting to change the landscape of the industry by investing heavily in new plant builds. According to industry sources, more than 90 companies have announced plans for new polysilicon plants. Although we believe that many (perhaps most of these) will never have a material degree of success at this difficult process, even the expansion plans announced by well-tooled incumbents and second-tier players with strong expertise in chemical processes are sufficient to reshape the industry in the coming years and potentially lead to more modules than can be absorbed at current price levels, in our view (please refer to Exhibits 26 and 27).




EXHIBIT 26: PRIMARY SILICON MANUFACTURERS AND GROWTH PLANS

Announced Global Polysilicon Capacity Additions

5,000 10,000 15,000 20,000 25,000 30,000

Others

Sumitomo

Mitsubishi

MEMC

Tokuyama

REC

Wacker

Hemlock

(Metric Tons)2005 2006 2007 2008 2009 2010

Source: Hemlock Semiconductor, excludes UMG silicon


0



EXHIBIT 27: SILICON SUPPLY MODEL (METRIC TONS)

Tier Company 2006 2007 2008E 2009E 2010EEstablished Hemlock Semiconductor Corp. 10,000 10,000 14,500 19,000 27,500

Wacker-Chemie 6,500 7,500 10,000 12,575 16,900REC 5,500 5,500 6,900 12,500 19,500MEMC Electronic Materials 4,000 6,200 7,000 11,500 15,000Tokuyama 5,200 5,200 5,300 7,800 7,800Mitsubishi (Materials & Polysilicon) 2,650 2,650 3,200 3,200 4,300Sumitomo 750 750 1,000 1,300 1,300

Sub-total 34,600 37,800 47,900 67,875 92,300% of overall supply 94% 90% 78% 65% 54%

Emerging DeGussa/SolarWorld 100 200 850 850PV Crystalox 900 1,100Nitol 1,878 3,515DC Chemical 1,800 5,500 13,000M Setek 600 2,800 3,542 6,240

Sub-total 700 4,800 12,670 24,705% of overall supply 2% 8% 12% 15%

New entrants LDK Solar 150 3,200 6,900Hoku Scientific 292 2,800Asia Silicon 200 1,500 2,000Sichuan Xinguang 300 500 1,000 2,000Luoyang Zhonggui 300 1,200 2,000 3,000Jiangsu Zhongneg (GCL) 300 1,300 1,950 7,200Emei Semiconductor 200 200 600 700 2,200AE Polysilicon 200 1,200 2,550Renesola 300 2,400SILFAB 1,250All other Chinese companies ignored

Sub-total 1,100 4,150 12,142 32,300% of overall supply 3% 7% 12% 19%

Metallurgical Elkem 2,000 4,000Timminco 1,098 5,000 9,200Globe Specialty Metals 360 1,000 1,000

Sub-total 0 1,458 8,000 14,200% of overall supply 0% 2% 8% 8%

Other 2,100 2,500 3,168 4,489 6,104

Total Silicon Production (m tons) 36,700 42,100 61,500 105,200 169,600Source: Jefferies’ estimates, Hemlock Semiconductor, Wacker-Chemie, REC, MEMC, LDK, Bloomberg, Reuters. All figures are estimates.




Upgraded Metallurgical (UMG) Silicon

Alternative methodologies of silicon processing to reach purity levels sufficient for PV usage have gained increasing levels of attention in recent years. Metallurgical silicon (m-Si) is a commodity that is currently used in the traditional Siemens reactor silicon production process. In this process, m-Si is gasified and is used as part of the trichlorosilane gas, which is injected into the deposition furnaces. Under the process developed by Elkem, Timminco, and several others, the raw m-Si is not gasified but instead purified directly into solar grade silicon. The benefit is the elimination of the gasification step which could result in an end product whose production cost is 30%–50% cheaper than silicon produced in the traditional Siemens-reactor process, and capacity could be brought online much faster than for a traditional Siemens process.

Historically, the UMG process itself was too expensive and yielded cells with conversion efficiency and quality levels that were inadequate for PV usage. However, recent advances in the economics of the process, as well as learning curve experience working with UMG wafers on pilot lines, has allowed certain manufacturers to begin selling PV cells made with UMG silicon, raising the question of whether or not tight silicon availability will remain a risk in future cycles going forward. Cell manufacturers such as Q-Cells and Canadian Solar are reporting good progress using UMG wafers from Elkem and Timminco to make cells with conversion efficiency levels as high as the mid-teens for 100% UMG. Yet, we note that reported yield levels from UMG factories are currently very low compared with normal polysilicon, and that UMG wafers currently cannot be used to manufacture higher range conversion efficiency cells. Since small changes in conversion efficiency can drive significant changes to gross margins, switching from normal polysilicon to UMG would only make economic sense if the presumed loss in gross margin (due to a drop in conversion efficiency) for an installed PV system is more than offset by the cost reduction of UMG versus polysilicon. Additionally, we are concerned that the low conversion efficiency and lower purity of UMG could lead to higher conversion (transforming raw silicon to wafer to cell to module) costs. Finally, lower efficiency modules will likely be sold at a discount to traditional, higher efficiency modules. With polysilicon prices set to come down with increased expansion, we remain somewhat cautious on the long-term cost-benefit proposition of UMG.

Pricing Impact

Given the potential for additional silicon supplies to hit the market in the coming months/years and the current uncertainty around growth in the size of the various incentive pools available for solar, we project that PV system prices will need to fall faster than the build-in rate of digression in key markets like Germany to create sufficient demand to absorb the supply of modules coming on. Given the lack of accurate aggregate data available for silicon production being sold to PV cell and module manufacturers, as well as the difficulty of measuring price elasticity of demand in this relatively new and highly subsidy-driven market, determining precise inflection points for silicon or module availability and price changes is difficult. Our European and U.S. solar analysts agree that module ASPs could fall 15-20% in 2009 from current levels.

Cost Reduction Opportunities

In most PV markets, incentives are expected to decline over time to “force-wean” the solar industry off subsidies that were meant to encourage scale economics. Thus in order to keep adoption rates high through attractive equity rates of return and pay-back periods, installed solar costs (not just modules) will need to drop at least sufficiently to offset the decline in incentives. In markets like Germany, where feed-in tariffs are structured on a declining scale, this allows strong visibility into the need to cut installed system prices. Yet in practice, gyrations in demand resulting from the introduction of new incentives in other (competing) markets can create the opposite effect, with system prices rising as subsidies fall.

Against a backdrop of declining system prices, PV manufacturers are targeting aggressive cost reduction opportunities related to scale, silicon costs, conversion efficiency levels and vertical integration. As the industry grows, the opportunity to source materials and equipment from multiple suppliers in larger volumes also grows, allowing for scale benefit opportunities. In Exhibit 28 below, we reflect SunPower’s cost reduction plans of at least 50% by 2012. We note the prominence of downstream (installation and distribution) savings in the overall mix, which we believe may be among the most difficult to accomplish without more vertical integration and consolidation of market share in that segment.




EXHIBIT 28: COST REDUCTION TARGETS

Source: SunPower Presentation

Conversion Efficiency

Increased cell efficiency is a very powerful driver of cost reduction for PV systems, based on the simple premise that each increase in conversion efficiency yields more watts per every unit of input (silicon, labor, power, capex, etc.) throughout the value chain. Translate this into more kWh output per unit of cost and it is easy to see why pursuing higher conversion efficiency is a profitable exercise. For example, a call and module manufacturer able to raise module conversion efficiencies by 100 bps from 14% to 15%, raises module output by about 7%, which in a static world translates into a 7% price premium for the module and an almost 600 bps increase in gross margin (assuming gross margin prior to the change was about 25%). Since the difference in module efficiency levels between the best and average industry performers is currently as much as 700 bps, adoption of best practices could have a substantial impact on industry cost structures.

Economies of Scale

The solar industry has ample room to benefit from increasing economies of scale. It is important to note that, given the small size of the industry, individual companies have not yet had a chance to grow to optimum size. Case in point, the three largest listed European solar companies (SolarWorld, REC, & Q-Cells) had combined revenues of less than €2.4 billion in 2007. We believe the growth in the industry will lead to significant improvements through higher automation, more straight-through processing, and improved production yields. This process is at its early stages and we should see the benefits in the coming quarters and years as operating costs as a percentage of revenue should show stable to positive trends despite falling unit prices. Solarworld estimates that for every doubling of its production, per unit costs have declined by 20%.

Silicon Efficiency

Crystalline silicon solar cells represent about 90% of the total solar market and rely on silicon as the key feedstock. The surging demand for solar cells has led to a jump in silicon prices. As recently as 2005, solar companies could secure feedstock at $30–$50/kg, while current new silicon supply contracts are being priced at levels often above $100/kg and many have pricing mechanisms with links to moves in a basket of spot prices (although not priced at spot-type rates). On a spot basis, the price for silicon has exceeded $500/kg for high-quality silicon that is blended with lower-quality scrap, although it is unlikely that these prices could be economical for any cell producer not attempting to blend.

In this context, cell and module producers look for ways to maximize their output from by increasing conversion efficiency levels as discussed above, by reducing the thickness of wafers and by reducing losses from the production process by reducing “kerf loss” and improving the processes of recycling any scrap materials.

Thinner Wafers

Reducing wafer thickness can lead to a higher volume of wafer production from a static amount of silicon. However, a reduction in wafer thickness does not lead to a linear increase in wafer production unless kerf, or cutting in layman’s terms, losses can be reduced as well. In the Thinner Wafers scenario shown in Exhibit 30, we




assume that wafer thickness is reduced 37% from 240 microns to 175 microns but the silicon cost per Wp (Watt peak) falls only 17% due to constant wire thickness.

EXHIBIT 29: WAFER THICKNESS

Development in wafer thickness (µm)

325

280

240

200

160

100

150

200

250

300

350

2003 2004 2005 2006 Future

9% more wafers per silicon unit

19% more

31% more

46% more

Source: REC Capital Markets Day Presentation

EXHIBIT 30: SILICON SAVINGS

Base Scenario1 Kg Silicon = 29 Wafers (@ 240 microns) x 29 Cells w/ 3.6Wp Output (@ 15% cell effiency)

1 = x 3.6 = 104.4 Wp Production€ 40 = € 0.39 Silicon cost/Wp

Thinner Wafers1 Kg of silicon = 35 Wafers (@ 175 microns) x 35 Cells w/ 3.6Wp Output (@ 15% cell effiency)

1 = x 3.6 = 126 Wp Production€ 40 = € 0.32 Silicon cost/Wp

Cost Savings 17%

Higher Cell Efficiency1 Kg of silicon = 29 Wafers (@ 240 microns) x 29 Cells w/ 4.1 Wp Output (@ 17% cell efficiency)

1 = x 4.1 = 118.9 Silicon cost/Wp€ 40 = € 0.34 Silicon cost/Wp

Cost Savings 12%

Thinner Wafers & Higher Cell Efficiency1 Kg of silicon = 35 Wafers (@ 175 microns) x 29 Cells w/ 4.1 Wp Output (@ 17% cell efficiency)

1 = x 4.1 = 143.5 Silicon cost/Wp€ 40 = € 0.28 Silicon cost/Wp

Cost Savings 27%

35

35

29

29


Given that PV cell and module producers could see significant reductions in pricing if incentives do not expand rapidly, we have analyzed a few basic cost reduction opportunities in the context of rapid ASP declines. In the first scenario in Exhibit 31 below, we ask the question: “How much could ASPs decline while keeping margins flat?” Implicit in our assumption is that cell and module producers will be able to raise conversion efficiency levels and benefit from cost reductions related to the increase in the amount of available silicon combined with a move to more long-term fixed silicon contracts. For several Chinese companies that have very low costs of processing cells and modules, but very high costs of silicon due to heavy reliance on the spot market, the results are encouraging. We estimate in the scenario below that raising module conversion efficiency levels from 13% to 14% and seeing a




17.5% reduction in silicon prices could allow margins to stay constant with a 19% drop in ASPs. We note that module efficiencies are typically lower than cell conversion efficiencies by perhaps 200 bps.

EXHIBIT 31: HOW MUCH COULD ASPS DECLINE WHILE KEEPING MARGINS FLAT? (ROUGH ESTIMATES)

Base Case (margins fixed) Curr 4Q08 2009 2010 2011 2012ASP $4.00 $3.90 $3.31 $2.68 $2.23 $1.95Change in ASP % -2% -15% -19% -17% -13%Cumulative Change in ASPs -2% -18% -37% -53% -66%Module cost/Watt $3.00 $2.93 $2.48 $2.01 $1.68 $1.46Gross margin 25% 25% 25% 25% 25% 25%

Avg Si wafer cost/kg $225 $225 $186 $139 $104 $78Chg in Si cost % 0.0% -17.5% -25.0% -25.0% -25.0%Wafer thickness (microns) 200 185 175 165 155 145Kerf loss (microns) 150 150 145 140 135 130Conversion efficiency 13.00% 13.00% 13.50% 14.25% 15.00% 15.50%Grams/Watt 7.7 7.4 6.8 6.1 5.5 5.1Si wafer cost/Watt $1.73 $1.66 $1.26 $0.85 $0.58 $0.40Non-Si cost/Watt $1.27 $1.27 $1.22 $1.16 $1.10 $1.06Module cost/Watt $3.00 $2.93 $2.48 $2.01 $1.68 $1.46Si cost/total module cost 58% 57% 51% 42% 34% 27%

Cost reduction breakdown %Silicon prices 0% 60% 60% 58% 61%Wafer thickness & kerf loss 96% 4% 3% 3% 3%Efficiency 0% 24% 28% 30% 25%Other 4% 12% 9% 9% 10%Total 100% 100% 100% 100% 100%

Source: Jefferies & Company, Inc. estimates; ASP declines driven by cost declines and assumed fixed margins

Yet, what if incentive levels are sufficient to maintain a higher level of pricing (or less of a drop)? What could margins grow to in our scenario? In Exhibit 32, we estimate ASP declines of 5% in 2008, followed by 10% p.a. in the following years, leading to gross margin gains.

EXHIBIT 32: WHAT IF ASPS DON’T DECLINE BY THAT MUCH?

Base Case (margins adapt) Curr 4Q08 2009 2010 2011 2012ASP $4.00 $3.90 $3.51 $3.16 $2.84 $2.56Change in ASP % -2% -10% -10% -10% -10%Cumulative Change in ASPs -2% -12% -22% -32% -42%Module cost/Watt $3.00 $2.92 $2.46 $1.98 $1.64 $1.42Gross margin 25% 25% 30% 37% 42% 45%

Avg Si cost/kg $225 $225 $186 $139 $104 $78Chg in Si cost % 0.0% -17.5% -25.0% -25.0% -25.0%Wafer thickness (microns) 200 185 175 165 155 145Kerf loss (microns) 150 150 145 140 135 130Conversion efficiency 13.00% 13.00% 13.50% 14.25% 15.00% 15.50%Grams/Watt 8.0 7.7 7.0 6.4 5.7 5.3Si cost/Watt $1.80 $1.72 $1.31 $0.89 $0.60 $0.41Non-Si cost/Watt $1.20 $1.20 $1.16 $1.09 $1.04 $1.01Module cost/Watt $3.00 $2.92 $2.46 $1.98 $1.64 $1.42Si cost/total module cost 60% 59% 53% 45% 37% 29%

Cost reduction breakdown %Silicon prices 0% 60% 61% 59% 62%Wafer thickness & kerf loss 96% 4% 3% 3% 3%Efficiency 0% 24% 27% 29% 24%Other 4% 12% 9% 9% 10%Total 100% 100% 100% 100% 100%

Source: Jefferies & Company, Inc. estimates; margin changes driven by ASP declines and assumed cost reduction opportunities




Please note that the examples above are intended to illustrate how cost savings would play an important part in maintaining margins in a deflationary environment. The pricing forecasts used are for illustrative purposes only and do not reflect any specific company, nor the analysts’ views on the exact timing of ASP falls in the near term.

Solar Economics

Without subsidies, solar PV is generally not competitive today with metered electricity prices in most markets. Thus, subsidies are the key driver of the economics of solar installations. Excluding subsidies, solar economics are driven by the cost of the installed system, the amount of sunlight and the cost of grid priced electricity that one is avoiding by using solar energy (or the next best generation option in off-grid on non-distributed generation scenarios). Today, there is only a weak correlation with solar adoption and the amount of sunshine in certain regions. Some of the best solar markets are in climates with relatively poor radiation characteristics are (Germany and Japan), although new robust markets have developed around strong subsidies in sunny regions like California, Spain and more recently Italy where solar conditions contribute to better economics.

EXHIBIT 33: SOLAR RADIATION

Source: NASA & REC Annual Report

Several areas of Mediterranean Europe would appear to have excellent solar characteristics based on high electricity costs as well as abundant sunshine as depicted in Exhibit 33. We note though, that other factors such as the amount of available land and rooftops to dedicate to solar installations are also factors. Naturally, in the near term, subsidies have the strongest impact on solar adoption.

EXHIBIT 34: UNSUBSIDIZED PV GRID PARITY RATES IN EUROPE BY AVERAGE INSOLATION LEVELS (ESTIMATES)

Source: Applied Materials analyst day presentation January 2008




Cost of Production Is Inappropriate Comparison

In Exhibit 35 below we depict estimated generation costs per MWh for newly build plants for traditional forms of generation as well as wind and solar, assuming “fully loaded” emissions costs for NOx, SOx, mercury and CO2 (at $25/MT). While solar has by far the most expensive generation cost of the group, we note that when solar is being used as a distributed form of generation (e.g., rooftops), its generation cost should be compared with the price of power at the meter, not generation costs of electricity generated from other technologies that must then be transmitted through the grid. Yet, even on this basis today, solar generation can look expensive.

When comparing the viability of solar power costs on rooftops (i.e., distributed generation), we think it is important to compare the cost of solar production with the retail cost of energy at the meter for which the power from solar modules is mean to be a substitute, rather than to the cost of power generation. Today, unsubsidized solar generation costs of $0.25–$0.50/KWh are no match for average retail rates of around $0.10/KWh in the U.S. (generally close to double that level in much of Europe). Yet, in many markets grid prices are substantially higher ($0.15–0.30/MWh), which should allow solar to more effectively close the gap in certain regions. For utility scale solar farms, we believe that the comparison with generation costs may be more appropriate, given the needs for transmission to reach the point where the KWh are used.

Obviously, some of these assumptions are not indicative of the current environment for emissions in the US, as CO2 is not under mandatory restrictions in most coal generating regions and mercury regulations have not yet kicked in (or are the costs of compliance clear). However, we are trying to create a view on the regulatory scenario we believe new plant builds will face in the coming years. We admit that some assumptions are also over-simplifications, such as coal plants needing to bear the full cost of NOx and SOx reductions through the auction market (rather than with scrubbers and SCRs) and no blending of coal to lower all-in fuel costs. However, even with these modifications, we believe the rapid run in coal, oil and natural gas prices have already begun having a significant effect on baseload and peaking generation costs and assessments of investment decisions about different generation technologies. Moreover, the price of carbon abatement can have a significant impact on relative generation costs between coal and natural gas fired plants compared with wind. We also note that we have not included the benefit of REC (renewable energy certificate) values in these calculations for solar and wind.

EXHIBIT 35: GENERATION COST COMPARISON

Coal Nat Gas Nuclear Wind SolarCapex/kW $2,000 $900 $2,500 $2,000 $6,500Fuel/unit $100 $8 $65 NA NASWU price NA NA $148 NA NABtu/unit 12,500 1,000 NM NA NAHeat rate 10,000 7,000 NM NA NASO2/ton $150 $0 $0 $0 $0NOx/ton $800 $0 $0 $0 $0Hg/lb $30,000 $0 $0 $0 $0CO2/MT $25 $25 $0 $0 $0Cost per kWhFuel & other $40.00 $53.62 $9.50 $0.00 $0.00D&A $9.51 $4.28 $11.89 $27.67 $140.00Financing $7.14 $3.21 $8.93 $25.18 $204.75O&M $9.00 $3.50 $10.51 $7.50 $5.00SO2 $0.60 $0.00 $0.00 $0.00 $0.00NOx $1.40 $0.00 $0.00 $0.00 $0.00Hg $1.50 $0.00 $0.00 $0.00 $0.00CO2 $23.30 $9.94 $0.00 $0.00 $0.00Subsidies $0.00 $0.00 $0.00 ($20.00) ($84.61)Total per kWh cost $92.45 $74.56 $40.83 $40.36 $265.14

- - - - 60.00

0

50

100

150

200

250

300

350

400

$/MWh

Coal Nat Gas Nuclear Wind Solar MeteredPrice

New Generation Facility Cost Comparisons

Subsidies

Emmissions costs

Max

Min

Source: Bloomberg, Platts, uxc.com, Jefferies & Company, Inc. estimates

Pay-back Periods and IRRs

While a certain portion of solar adopters undoubtedly have “green” motivations, we believe that most consumersand investors driving solar adoption rates are responding rationally to market forces and seeking attractive rates ofreturn and pay-back periods. In our view, when pay-back periods extend beyond the length of time a typical familyowns a home, residential solar is a tough sell. When commercial PV installations cannot match the returnsavailable from other projects, we believe their adoption rates also suffer. In additional to attractive subsidies, thehistorically low interest rate climate has allowed investors to highly lever installed costs (net subsidies), drivingstrong equity returns and paybacks, even when overall project returns may be lackluster. Naturally, this means thatsolar installations should be very sensitive to interest rates, a theory not yet tested by the expansionary monetaryconditions we have experienced in recent years during the development of the solar industry.




In Exhibit 36 below, we estimate pay-back periods for unsubsidized versus 50% subsidized PV systems at different rates of electricity, using an assumption of 1,500 hours of solar-harnessed sunshine a year (i.e., a very sunny climate) and no financing costs. The unsurprising result is that the cost of installed PV systems must get much cheaper than the current $7–$9/watt for rooftop systems and/or electricity prices must get much more expensive for PV to see rapid unsubsidized adoption. In many situations, PV would have no meaningful unsubsidized pay-back period given that modules have an estimated life in the environs of 25 years.

EXHIBIT 36: PAY-BACK PERIODS IN YEARS (ROUGH ESTIMATES)

$9.00 $8.00 $7.00 $6.00 $5.00 $4.00 $3.00 $2.00 $1.00$0.100 60.0 53.3 46.7 40.0 33.3 26.7 20.0 13.3 6.7$0.125 48.0 42.7 37.3 32.0 26.7 21.3 16.0 10.7 5.3$0.150 40.0 35.6 31.1 26.7 22.2 17.8 13.3 8.9 4.4$0.175 34.3 30.5 26.7 22.9 19.0 15.2 11.4 7.6 3.8$0.200 30.0 26.7 23.3 20.0 16.7 13.3 10.0 6.7 3.3$0.225 26.7 23.7 20.7 17.8 14.8 11.9 8.9 5.9 3.0$0.250 24.0 21.3 18.7 16.0 13.3 10.7 8.0 5.3 2.7$0.275 21.8 19.4 17.0 14.5 12.1 9.7 7.3 4.8 2.4$0.300 20.0 17.8 15.6 13.3 11.1 8.9 6.7 4.4 2.2$0.325 18.5 16.4 14.4 12.3 10.3 8.2 6.2 4.1 2.1$0.350 17.1 15.2 13.3 11.4 9.5 7.6 5.7 3.8 1.9$0.375 16.0 14.2 12.4 10.7 8.9 7.1 5.3 3.6 1.8$0.400 15.0 13.3 11.7 10.0 8.3 6.7 5.0 3.3 1.7

$9.00 $8.00 $7.00 $6.00 $5.00 $4.00 $3.00 $2.00 $1.00$0.100 30.0 26.7 23.3 20.0 16.7 13.3 10.0 6.7 3.3$0.125 24.0 21.3 18.7 16.0 13.3 10.7 8.0 5.3 2.7$0.150 20.0 17.8 15.6 13.3 11.1 8.9 6.7 4.4 2.2$0.175 17.1 15.2 13.3 11.4 9.5 7.6 5.7 3.8 1.9$0.200 15.0 13.3 11.7 10.0 8.3 6.7 5.0 3.3 1.7$0.225 13.3 11.9 10.4 8.9 7.4 5.9 4.4 3.0 1.5$0.250 12.0 10.7 9.3 8.0 6.7 5.3 4.0 2.7 1.3$0.275 10.9 9.7 8.5 7.3 6.1 4.8 3.6 2.4 1.2$0.300 10.0 8.9 7.8 6.7 5.6 4.4 3.3 2.2 1.1$0.325 9.2 8.2 7.2 6.2 5.1 4.1 3.1 2.1 1.0$0.350 8.6 7.6 6.7 5.7 4.8 3.8 2.9 1.9 1.0$0.375 8.0 7.1 6.2 5.3 4.4 3.6 2.7 1.8 0.9$0.400 7.5 6.7 5.8 5.0 4.2 3.3 2.5 1.7 0.8

Installed cost per Watt (Unsubsidized)

Met

ered

KWh

pric

e

Installed cost per Watt (Subsidized ~50%)

Met

ered

KWh

pric

e

Source: Jefferies & Co. estimates, calculated assuming 100% of KWh produced is used by module owner in lieu of grid KWh, excluded financing costs

In Exhibit 37, we estimate the pay-back period and return for a German residential adopter receiving a FIT of€0.47/kWh for 20 years. While the upfront cost of the system makes the overall project pay-back and returnlackluster, if the owner is able to secure low-cost financing for the bulk of the system, the out-of-pocket investmentis reasonable and the returns on equity and pay-back on that initial investment are quite high.




EXHIBIT 37: GERMAN RESIDENTIAL PV EXAMPLE

EU Residential Installation (Germany/Rooftop)System size (Watts) 3,500 % leverage 90%Cost per watt 5.00 € Financed amount 15,750 €FIT 0.420 € Equity contribution 1,750 €Tax rate 35.0% Debt rate 5.5%System cost 17,500 € Equity hurdle 8.0%less: PV of FITs (18,987 €) WAAC 5.8%Net system cost (1,487 €) IRR (unlevered project return) 6.8%Total KWh generated over system life 93,510 IRR (levered equity return) 44.3%System life in yrs 25.0 NPV 1,406 €Grid price (hurdle) 0.15 € Payback period ~10 yrsNet power cost/KWh* 0.06 € Equity payback ~1 yr

* Assumes no credit for PV KWh generated beyond amounts usedSource: Jefferies & Company, Inc.

In the US, commercial incentives are quite attractive when the combined impact of the federal ITC, state level subsidies, accelerated depreciation and other factors are considered. In Exhibit 38 below, we estimate the returns and pay-back on a commercial system in California. Here both project and equity returns and pay-back periods are compelling.

EXHIBIT 38: U.S. COMMERCIAL EXAMPLE (CALIFORNIA)

US Commercial Installation (California)System size (Watts) 30,000 % leverage 80%Cost per watt $6.25 Financed amount $150,000Federal tax credit 30% Equity contribution $37,500Special depreciation bonus 50% Debt rate 6.3%Tax rate - Federal 35.0% Equity hurdle 12.0%Tax rate - State 6.0% WAAC 7.4%State PBI/KWh $0.22 System life in yrs 25.0

System cost $187,500 Grid price (hurdle)/KWh $0.20less: Tax credit @ 30% ($56,250) Rate of grid price increase 3.0%less: PV of state PBI ($34,743) Net Power Cost per KWh $0.13less: PV of net depreciation tax benefit ($12,934) IRR (unlevered project return) 9.0%less: PV of avoided costs, net tax impact ($65,608) IRR (levered equity return) 169.8%plus: PV of debt payments, net tax impact $100,848 NPV $11,708Net cost $118,813 Payback period ~6-7 yrsTotal KWh generated over system life 902,953 Equity payback ~1 yr

Source: Jefferies & Company, Inc.

Grid Parity – What It Isn’t

Grid parity refers to the point at which the cost of energy generated from an installed system is able to compete with the metered price of electricity. We expect sunnier markets with high electricity costs (Hawaii, Italy) to reach the point of grid parity earlier than others. We suspect that a target of around 2012 is a reasonable one for when solar could become competitive on a broad scale in attractive solar markets, based on cost-reduction opportunities and a view of rising metered electricity prices.

Yet we disagree with the suggestion that grid parity alone is sufficient to drive PV adoption at a pace that would satisfy financial investors in the space. If grid parity simply means an imbedded cost of energy on a per KWh-basis that is the equivalent of the grid, we point out that this hurdle can easily be met while still incurring very long pay-back times and low or negative IRRs. Fortunately, we also expect that solar subsidies are likely to be maintained beyond the point of grid parity, continuing to drive rapid adoption for some time. One caveat: if grid parity can be reached including the costs of financing the installation, we believe that this will be sufficient to drive rapid adoption as the combination of a low upfront investment and greater certainty on long-term energy costs are an attractive sales point.




The Solar Value Chain – Capturing Returns

EXHIBIT 39: SOLAR PRODUCTION CHAIN AND KEY PARTICIPANTS

Solar Value Chain ParticipantsSilicon Ingot / Wafer Cell Module

Hemlock M. Setek Q-Cells Solon

Wacker JFE Motech MSK

Tokuyama Renesola Aleo Solar

MEMC Sumco Centrosolar

DC Chemical

Will enter in 2009 PV Crystalox

Will enter in 2008 SolarWorld

Renewable Energy Corporation

Reducing exposure to wafering Sharp

ErSol

BP Solar

SunPower

Suntech

Sanyo

IsofotonTotal

Estimated 11-15 16-25 30-40 400+Participants

Source: Q-Cells Analyst Presentation

As indicated in Exhibit 40, silicon producers have the highest margins in the industry, followed by wafer manufacturers then cell and module makers. We note that capital investment tends to follow the bottleneck, although some skill sets, capital and equipment hurdles can extend the time necessary to “de-bottleneck” some segments of the value chain, and implicitly drive down returns. We believe that this is what we have seen in the silicon industry over the last few years due to the scarcity of talent and equipment necessary as well as the time and concentration of capital necessary to effectively construct a new silicon plant. As capital flows shift through the value chain, we believe that the next step may be further up the chain again — equipment manufacturing. Here recent private equity investments and interest from well-established public companies may be a sign of economic rent to be had. In other areas of the value chain, sustainable strategic cost advantages are rare (e.g., First Solar) and we believe that many business models will eventually undergo (or are already undergoing) real and virtual vertical integration to sustain margins as pressures grow to reduce overall PV system costs to grid parity.




EXHIBIT 40: VALUE CAPTURE IN THE PRODUCTION CHAIN

Volumes ASP/ UnitRevs $MM

EBIT % Capex/ Unit Cost $MM

Silicon

2,000 MT $160.0 32-44%

Wafers

250 MW $500.0 22-30%

Cells

250 MW $750.0 17-25%

Modules

250 MW $1,000.0 5-10%

250 MW $1,000.0 5-10%

Comments

Most costs volume based, so higher price

= higher margin

Thinner wafers & saws can increase wafer/kg

output

Higher efficiency cells can increase cell/wafer

output

$300.0

$0.50-0.90/Watt $175.0

$0.25-0.75/Watt

$125.0

Raw s ilicon with a purity of 90% is melted at around 1900°C and treated with siloxane gas, oxygen, and hydrochloric acid to raise purity to 99.9999%. Resulting product is chunks or granulated pure s ilicon. Approximately 1/3 the cost of production is energy usage.

Raw silicon is melted in a crucible and then poured into a mold to form an ingot which is subsequently sectioned and sawed into a poly-crystalline wafer (see below for diagram). Mono-crystalline ingots are created using a seed crystal and drawing the ingot from the silicon melt before being squared off and sawed into wafers. Approximately 40% of silicon is lost in the sawing process

Solar wafers are first acid washed to erase sawing damage and to create a “dimpled” surface to maximize absorption. The wafer then undergoes a series of chemical and thermal treatments to introduce an electrical field and reduce reflectivity before a metallic paste is screen printed on to the cell to provide electrical contacts. Current poly-crystalline cells have approximately 15% conversion efficiency (mono-crystalline cells are about 200 bp higher) and generate around 4 Wp per 156mmx156mm cell.

Solar cells are connected in an aluminum frame and sandwiched between protective glass to shield the delicate cells. Note that module efficiency is lower (generally 100–200 bp) than the efficiency of the component cells as spacing between the cells is required to prevent a short. Sometimes, modules are mounted atop movable tracking systems that maximize sun exposure.

$45/kg legacy contracts ->

500+/kg low vol spot mkt

~$3.50-$5.50 per 156mm x 156mm

wafer, includes cost of silicon

~$2.75-$3.25 per Watt

$0.15-0.25/Watt

Higher efficiency cells increase watts/module

output

~150k/MT

$0.15-0.25/Watt

$50.0

Installation / Downstream

~$3.50-$4.25 per Watt

Solar modules are grouped into an array and mounted on rooftops using brackets that are attached to the roofing structure. The panel is connected to an inverter which converts the power to AC in order to be fed into the grid. Total cost of installation can be ~$2-4/watt including the inverter, materials, labor, and installer profits.

~$3.50-$4.25 per Watt

Higher efficiency cells increase watts/module

output

$50.0

Source: Jefferies’ Estimates

Thin Film – Disruptive Technology or New Market Opportunity?

Thin film PV refers to products that do not use traditional crystalline silicon wafers to produce the photovoltaiceffect (and therefore are not subject to tight silicon availability). Thin film is considered by many to be a potentiallydisruptive technology in the solar industry, due to the ability to use higher throughput, less labor intensivemanufacturing techniques to create less expensive PV products. Thin film typically uses about 1% of thesemiconductor material necessary to generate the same amount of power from crystalline silicon-based products.Moreover, some thin film products may be suited for applications that traditional crystalline silicon modules may notbe able to serve, such as those requiring flexible, lightweight materials, which we believe could be important foradvanced BIPV (building integrated photovoltaic) solutions, such as imbedding PV materials in building materials.Today lower conversion efficiency levels for thin film can work against their superior manufacturing costs for certainapplications, yet the high cost of silicon for crystalline silicon producers has increased the gap between thin filmand crystalline silicon module manufacturing costs, creating fertile opportunity for thin film to grow faster than thePV market overall. Conversion efficiencies ranges from approximately 6% to 11% for commercial thin film cellscompared with 14%–22% for crystalline silicon cells. Solarbuzz estimates that thin film accounted forapproximately 12% of the PV product produced in 2007, up from 8% in 2006, although other industry consultantsplace the percentage much lower, given their higher estimate for total PV production.

Amorphous Silicon (A-Si) – Despite the name, amorphous silicon uses no polysilicon and is not subject to supplyconstraints in that market, but does use small amounts of silane gas, which is one of the building blocks of thepolysilicon process. The technology is among the better understood versions of thin film technology and is being




manufactured commercially on both glass substrates modules and thin, flexible modules. Unisolar (Energy Conversion Devices [ENER, Buy]) is the largest flexible manufacturer of A-Si. Applied Materials (AMAT, NC) is rolling out an equipment suite for customers to begin manufacturing A-Si on glass in 2H08. A materials limitation described as the Staebler-Wronski effect causes A-Si materials to degrade when exposed to sunlight, theoretically limiting the potential conversion efficiency of the cells in the real world without developing multi-junction cells. Current claimed conversion efficiency levels for A-Si cells are approximately 8%–8.5% at Unisolar, while manufacturing costs are currently more than $2.00 per watt, although declining quickly with scale and better materials sourcing, among other cost reduction opportunities. Conversion efficiencies on a module basis as calculated by Jefferies & Company, Inc. from company product spec sheets suggests that on a surface area basis, efficiencies are lower due to the amount of surface area of the module not being used for generation. AMAT claims its SunFab suite will allow customers to produce A-Si modules at a cost of $1.50/watt.

Copper Indium Gallium Diselenide (CIGS) – CIGS hold the promise for the highest potential conversion efficiency among thin film and can be deposited on either flexible substrates (plastic or super thin stainless steel) or rigid (glass), although there is little commercial scale manufacturing of CIGS product currently. Lab tests have demonstrated conversion efficiencies on CIGS cells as high as 19.9%, although we believe that most CIGS manufacturers currently ramping or planning production will begin commercial operations closer to 10% conversion efficiencies. Initial attempts at CIGS production met with highly variable conversion efficiencies due to difficulty creating uniform cells when combining the CIGS materials. The promise of CIGS also lays in the potentially high speed industrial processes being used to manufacture the cells, including thermal co-evaporation, sputtering, and depositing nano-particles. Venture capitalists have invested heavily in private CIGS companies such as Nanosolar, Miasolé, Solopower, Solyndra, and Heliovolt. Public CIGS plays include Daystar Technologies (DSTI, NC) and Ascent Solar (ASTI, NC). However, those attempting to manufacture on flexible substrates have been experiencing significant issues related to encapsulating the modules in a manner that prevents degradation. Typical EVA (ethyl vinyl acetate) solutions that may be sufficient for other technologies have proven an insufficient barrier for CIGS applications that appear to be more susceptible to moisture. Thus, we believe that most initial scale commercial operations for flexible CIGS applications will be placed under glass for protection, eliminating the possible advantages of a flexible, lightweight solution. CIGS has been critiqued for potential materials sourcing issues with indium, which is rare and also used for the flat screen market (TVs, computer screens).

Cadmium Telluride (CdTe) – Arguably the most commercially successful thin film to date, CdTe modules are produced primarily by First Solar (FSLR, NC) at a highly competitive cost structure approaching $1/watt. The company has been able to replicate its production process very successfully, after some initial hiccups in its manufacturing processes. CdTe is generally only produced on glass, potentially limiting its long-term use as a rooftop application where weight is an issue, although First Solar has managed to sell many modules into rooftop applications. Conversion efficiency levels today are more than 10%, although they have shown potential in the lab to reach much higher levels and we believe that levels in the mid-teens should be achievable on a commercial scale. We note that First Solar’s use of a laser to separate the CdTe layer to create cells on its glass modules leaves little non-PV space on a module, leading to little difference between claimed and actual conversion efficiency levels as measured on a surface area basis (compared with other modules where claimed efficiency levels may differ considerably with surface-area efficiency. The use of cadmium as a primary raw material has created concerns about potential toxicity issues under certain conditions, although we know of no events where cadmium has been released in a toxic form from CdTe modules and we have not seen the modules barred from any markets or applications due to toxicity concerns. We note that First Solar incurs an accrual charge as a part of the cost of selling each module related to reclaiming the module at its useful life end. The use of tellurium, a very rare element, has raised questions about the ability to reach multi-gigawatt scale with CdTe, although First Solar has successfully secured substantial amounts of the material for future growth. The rapid throughput of First Solar’s production process allows for very favorable process economics that suggest that CdTe could be the initial choice of utility scale PV projects.




EXHIBIT 41: FIRST SOLAR TAKES A COMMANDING LEAD

Selected Thin Film Producers

0 50 100 150 200 250

Bankok Solar

Fuji Electric Systems

Mitsubishi Heavy Industries

Wurth Solar

Sinonar

Sharp

Trony

Kaneka

Unisolar

First Solar

2006 2007


EXHIBIT 42: C-SI TECHNOLOGY CONTINUES TO DOMINATE

Cell Market Shares by Technology

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1999 2000 2001 2002 2003 2004 2005 2006 2007

MulticrystallineMonocrystallineAmorphous SiliconThin Ribbon (C-Si)Cadmium TellurideCIGS/CIS


Other Alternatives

String Ribbon. String ribbon technology produces wafers by melting chunks or pellets of silicon in a crucible and“pulling” it out of the melt between ceramic strings, allowing it to cool as a thin layer or ribbon. The processobviates the need to create ingots and eliminates much of the silicon waste created through kerf loss. EvergreenSolar (ESLR, Buy) and Schott Solar are the primary manufacturers of this form of wafer that can use about half ofthe silicon used by other forms of wafer manufacturing (Evergreen uses about 5 grams per watt currently,




compared with more than 8 for the average cell producer, but expects this figure to eventually decline to 2.5 as yields and efficiency levels improve. String Ribbon can only be used to create multi-crystalline based cells and its efficiency levels are generally somewhat lower than average currently, although processes are being implemented to improve efficiency levels.

Gallium Arsenide. Gallium Arsenide solar cells have better absorption capabilities than traditional crystal solar cells and because of that characteristic demonstrate significantly higher conversion efficiency (~30% or more). The cells were developed for use in space applications prior to being adopted for potential terrestrial use with optical concentrators to focus sunlight on small versions of the cells. Without concentrators, the economics of terrestrial applications do not work, but with optics initial installation economics can be compelling. The major producers of Gallium Arsenide solar cells are Emcore Photovoltaics (EMKR, Buy) and Spectrolab.

Organic Solar Cells. At an earlier stage of development are organic photovoltaic receptors. These organic receptors act as semiconductors similar to silicon-based solar cells. The potential benefit for these cells is through a relatively simple manufacturing process in which organic cells can be printed onto a substrate, resulting in potentially lower costs. These organic cells have many of the same characteristics as the thin film cells today (lightweight, flexible). The disadvantage of organic solar cells are its relatively low efficiency of only 3%–5%, less than half the efficiency of current thin film products. Additionally, no companies are currently at commercial levels of production. The following companies are working on developing this technology: BP Solar, Global Photonic Energy Corporation, Konarka Technologies, Luna Innovations, and Nanosys.




Risks

The principal risks are associated with the high investment cost for solar installations versus traditional forms of generation but there are others as well. We have identified and provided our insight on some of the key risks below:

• Incentives. As we previously mentioned, government support is vital to sustain current growth expectations and accelerate cost reductions to the point where solar can compete on a stand-alone basis with traditional generation. We believe that a combination of energy security, energy prices, and increasing environmental awareness all will serve to ensure that government support grows rather than shrinks. However, the timing of changes to these incentives can create considerable volatility in share prices.

• Silicon availability and price. Lack of sufficient polysilicon production capacity to meet PV manufacturing needs has driven up prices of silicon and left some companies with excess manufacturing capacity, negatively affecting margins. Blending of high-grade polysilicon with scrap has begun to negatively affect or to slow down gains in conversion efficiency, a key cost driver. Mismatches between polysilicon plant investment cycles and near-term PV incentives can create significant bottlenecks.

• Interest rates. The genius of the feed-in tariff model is that it makes solar investment a financial rather than environmental decision. The downside to this is that if interest rates rise, the solar investment premium will decline and potentially reduce demand.

• Energy prices. Fluctuations in energy prices can have a significant impact on solar shares. While natural gas actually has a more relevant impact on electricity prices, it is oil price that seems to be a better proxy for public perception about energy prices. Oil can dictate short-term performance of solar shares due to the visibility of its impact on consumer decisions and a sense that politicians will feel compelled to pass pro-renewables legislation when oil prices are high. However, on a more fundamental basis, a sustained drop in energy prices would increase the time needed for solar to become cost competitive with traditional energy generation. However, we believe that high and rising energy prices will continue to contribute to the attractiveness of solar generation.

Conclusion

We believe that the solar sector represents an excellent long-term investment opportunity for investors, although we are aware that volatility will be high at times. The rising competition for fossil fuels, increased concerns over energy security, growing awareness of the impact of human activity on the climate, and the resulting spread in government incentive programs indicate to us that solar generation will be a growing and attractive arena for investors in the actual solar systems and also investors in the stocks of companies that manufacture the solar modules.








EventThis report is designed to provide investors with a background onthe development and future of the wind power industry.

Key Points• Low cost. Wind is the most affordable form of zero-emission

renewable energy and is cost competitive with traditional energyin some sites. On average, wind power costs approximatelyU.S. $0.07-0.09/kWh compared to a blended traditionalwholesale (coal, gas, nuclear) cost of $0.04-0.06/kWh. Atcurrent prices, wind is broadly competitive with natural gaspowered electricity and would be competitive with coal firedelectricity with a carbon cost of $25-$35.

• ... but costs (and prices) are rising as higher steel prices arebeing felt in rising turbine costs. Most turbine manufacturesattempt to pass these rising prices on to their wind farmdeveloper customers. The extent to which prices can continueto rise, before this starts to impact demand, remains to be seen.It should be noted that higher raw materials prices lead tohigher traditional energy costs as well.

• Improved technology allows turbines to capture more energyfrom the wind and improve reliability, which can lower the costof wind power produced despite higher turbine pricing.

• High scalability. Wind can easily be rolled out on a utility scalewith modern wind farm developments often reaching over 500MW. In 2007, more than 19.7 GW of wind capacity wasinstalled, which represented 14.8% of total new generationcapacity installed globally: Wind now represents 2.1% of thegeneration fleet and 1% of power production worldwide.

• New markets developing. Europe, led by attractive feed-intariffs in Germany and Spain, has traditionally been the keymarket for wind, representing around 60% of installations in the10-year period to 2007. The USA has begun to reach itspotential as the PTC has been in place since 2005 withoutinterruption. The USA represented 26% of installations in 2007and is likely to post excellent installations in both 2008E and,with a 12-month extension of the PTC in place, in 2009E. Chinaand India have emerged to claim seats at the top table in thewind industry. These two countries have represented more than20% of global installations in both 2006 and 2007.

• Wind does have limitations. In addition to the rising costs, themost significant issue is the mismatch between attractive windsites and load centers, which can inflate project costs. Also, bythe nature of their size, wind turbines can be very intrusive,which can create NIMBY (Not In My Back Yard) issues.

October 10, 2008

Clean TechnologyEnergy Generation - Wind

Clean TechnologyWind Power Primer

Investment SummaryThe wind energy sector has grown to become a significantproducer of zero emission power, and we believe wind continuesto show tremendous growth potential driven by rising energycosts and spreading pro-wind policies.


James Harris, Equity Analyst44 207 029 8691, [email protected]


Current State of the Wind Sector – Overview

Current themes in the sector, which are addressed in more detail in the main sections of this primer, are as follows:

Growth continues. The wind market continues to experience high levels of growth. In 2007, the wind industry posted record installation growth for the third consecutive year as annual installations reached 19,793MW while installed wind capacity has risen 29% CAGR over the past decade. Industry experts forecast future growth rates of over 25% p.a. for the foreseeable future.

Incentive-driven demand. The most significant driver of growth remains the existence of favorable wind incentives. Governments have continued to support pro-wind energy policies, as rising concerns over climate change and energy security have catalyzed the desire to emphasize development of zero-emission and sustainable sources of domestic energy generation. In Europe, feed-in tariffs were renewed and improved in Germany in June 2008. In the US, record installations were achieved in 2007 with the PTC in place. We expect another record year in 2008E as installers rushed to finalise projects before the original expiry date for the PTC in December 2008. As it turned out, the PTC was renewed for a further 12 month period until December 2009, as part of a wider renewable energy package bundled through Congress with the financial ‘bailout’ package. This should allow the US to maintain its position as a key wind market through 2009. We believe that the incoming Administration may look at longer term wind incentives as part of a wider energy policy review.

Utilities presence is growing. The trend towards larger-scale wind farms continued as utilities take an increasingly important role in the development of wind farms. This is a reflection not only of the recognition by some (but not all) utilities that wind is an increasingly common way of generating low-cost renewable energy, as well as the desire for turbine manufacturers to direct scarce turbines to well-financed customers at the expense of smaller, independent developers. The trend is exacerbated by the spread of Renewable Portfolio Standards (RPS) in many U.S. states which force utilities to secure a fixed percentage of their energy from renewable energy. This looks likely to continue, and the increasing scale of wind developments could put them beyond the scope of all but the largest independent wind farm developers.

MW size increasing. As a reflection of the trend towards utility scale wind deployment, the average size of turbines is also increasing. Larger MW machines provide a better energy yield per unit of investment and are therefore attractive for larger players. As a result, the 2-3 MW segment of the market continues to show above average growth. The next generation of turbines (over 5 MW), which will principally serve the off-shore markets of tomorrow, are still in the main at the development stage. With technology and cost issues still to be solved, the off-shore market is unlikely to really take shape until 2011 or beyond. In terms of the next steps for the technology, some of the leading turbine producers are now focusing on developing smaller, lighter, cheaper and more reliable 2-3 MW machines, rather than attempting to scale up to the >5 MW designs.

Rising steel costs. The significant increase in the price of steel and other raw materials has impacted the production costs for turbines, for which steel represents around 20% of the total cost. Some turbine producers have been able to pass-through these rising costs to their customers in the form of higher prices. However, steel costs remain a key area of concern for turbine manufacturers.

Supply bottlenecks. The turbine shortage continues and is likely to remain a limiting factor to industry growth through YE2008 with many producers suggesting that current order books are nearly filled until 2009. Bottlenecks in component supply continue to be main contributor to the shortage. Capacity is rising in the industry as component producers have increased confidence in the sustainability of growth forecasts. However, BTM Consult recently announced that they believe bearings will continue to be in tight supply for the next two years.

Turbine ASPs rising. With the steel cost pass-through and an ongoing turbine shortage, prices for wind turbines have increased significantly over the past few years. For example, BTM Consult suggest that 2007 saw an estimated 10% increase in the cost/MW of a turnkey wind farm (source – BTM Consult) although some turbine manufacturers posted greater increases in turbine sales prices. We believe that these increases have not yet had a noticeable impact on levels of demand.

Transmission capacity. . Current electricity grids are designed to transport power from existing power plants, often located near the relevant fossil fuel source and / or transportation hubs, to centers of demand. Wind farms, by their very nature, tend to be located relatively far from centers of demand and generally do not coincide with existing power plant sites which may lead to high transmission costs as new lines must be laid. Most wind farm develops seek to avoid this by ‘piggy-backing’ off existing transmission capacity, but capacity can be limited.

Energy Generation - Wind



Turbine Pricing – A Major Theme in Wind Power

Wind turbine prices, measured by price/capacity, have risen dramatically over the past five years in the principal European and US markets. In Europe, turbine prices have risen by approximately 25% while in the US the approximate increase has been a more dramatic 75% over the same period. There are several reasons for the rise in turbine prices:

� Rising steel and other raw material prices have a meaningful impact as steel represents approximately 20% of the cost of an assembled turbine.

� Higher transportation costs due to rising fuel prices and the emergence of new markets far from traditional production centers.

� Shift in pricing power from the developers towards the turbine manufacturers as demand has outstripped supply since 2005. This is particularly true in the US where developers tend to rush projects ahead of a potential lapsing of the PTC.

� Relative lack of an established production base in the US has forced the importation of components and even assembled nacelles, which not only increases transportation costs but also exposes USD prices to the impact of the stronger Euro.

However, we believe that the impact is not as significant as the headline figures may suggest and that returns on wind developments remain highly attractive despite the rising prices for several reasons:

� Prices have risen when measured in $ or € /MW capacity. However, turbine sizes have increased dramatically in the past five years. These newer and larger turbines are much more efficient at capturing energy from the wind which means more energy bang for the capex buck. Effectively, while the price per MW of capacity has increased, the price per MW hour of production has risen by far less.

� The three-year run of the PTC has allowed developers in the US to complete attractive, high load factor projects which allows the higher pricing to be offset by the better wind conditions. The flip side is that in some highly penetrated markets, demand can slow as flat feed-in tariff rate structures and a deterioration of project quality translates higher turbine pricing to lower returns. This is evident in Germany where annual installations are off 50% from the peak in 2002.

� Rising energy prices have increased base load electricity prices, which has improved returns for many projects, although wind farms in markets with a fixed feed in tariff structure do not benefit from this trend.

� Financing options have improved which has reduced debt costs for many projects. This rate of improvement is particularly strong in newer wind markets. The credit crisis has yet to play a major role in the growth of wind projects although clearly a weakened banking system could cast a shadow over the wind industry.

While the impact of higher turbine prices has not led to a global deterioration of wind project IRR in the past five years, it is worth noting that the desired trend for the renewable energy as a whole is lower costs. Although we are not concerned that wind may price itself out of existence in all but the most attractive wind sites and/or incentive regimes, we do note rising turbine price as a long-term risk for the wind sector.

Emergence of Large Scale Wind Developers

Another trend that has emerged over the past several years has been the increasing influence of large scale and/or utility wind developers. We believe this is a reflection of the increasing role of the utility scale wind farm, as well as the rising importance of the U.S. market where the PTC favors developers with existing tax bills. Further evidence of this trend can be seen in the increasing size of installed turbines as seen in the table below.

Turbine manufacturers also tend to favor the emergence of large, well-financed customers who can sign multi-year contracts covering hundreds of megawatts.

Independent wind farm developers and tax-driven private investors (very popular in Germany) are likely to decline in importance.

EXHIBIT 1: LEADING WIND DEVELOPERS

Developer MW InstalledIberdrola Spain 7,362Acciona Windpower Spain 5,300Florida Power & Light USA 5,077Electricidad de Portugal Portugal 3,696Babcock & Brown Wind Australia 1,859Long Yuan Electric Power China 1,620Eurus Energy Holding Japan 1,385NRG Energy Inc USA 1,300EDF Energies Nouvelles France 1,218Cielo Wind Power USA 1148Total 29,965

As a % of total installed capacity 31.8%

Source: BTM Consult




A Closer Look at Wind Turbines

The concept of harnessing the wind to provide energy has been around for centuries. Windmills first emerged in Persia in the 7th century and shortly thereafter appeared in both China and Europe. Early windmills clearly were not developed to reduce carbon emissions or offer peak shaving opportunities. Rather they were labor-saving devices designed to grind grain into flour and free up human and/or animal labor for more productive pursuits. However, in more recent times, the potential to use wind power to generate electricity has been noted and exploited.

− Superficially, there are few significant design differences between turbines produced by the major manufacturers. The most significant alternative solution is the gearless drive train offered by Enercon which offers higher reliability but at the cost of higher weight. Enercon also uses ultra-capacitors to drive blade pitch control which should, according to management, lower maintenance costs. In general, the product offering for the onshore market is fairly homogenous and producers have been focused on marginal reductions in operating costs through higher reliability and cost reduction in an attempt to differentiate their products.

− Reliability is a key consideration as turbine down time can trigger financial penalties to WTG producers and/or high maintenance costs to wind farm operators. The need for more robust machines is increased by the trend towards larger machines which, although they have a better MW per cost ratio, do create increased stress loads on the equipment. As a rule of thumb, energy output increases as a square of the blade diameter while the load on the drive train increases as a cubed function of blade diameter. Furthermore, larger rotor diameter increases asymmetrical stress on the WTG. For example, the force of the wind hitting the upturned blade is much higher than the force of the wind hitting the downward pointed blade(s).

− As a result of the increasing installation of utility scale wind farms, WTG manufacturers have also been asked to increase their grid compatibility to ensure a tighter integration with a grid principally designed to support traditional forms of generating capacity. This has forced producers to ensure that the wind developments ensure fault ride through capability (turbine does not trip from the grid during grid disruptions) and improved power control capability as the increasing weight of wind in the generation mix forces wind generators to perform certain grid stabilization operations normally carried out by traditional generators.

Anatomy of a wind turbine

As introduced above, modern wind turbines are built to a relatively homogeneous design, which is illustrated below.

EXHIBIT 2: STANDARD WIND TURBINE

Source: Winergy

With demand for turbines at record levels, the key for a turbine producer is to be able to manufacture the turbine efficiently and cost effectively. On the following pages, we discuss some of the key issues to consider for each component.




There are a number of themes to consider here:

• Value added. Some components (such as blades) are manufactured to company-specific specifications and are a key determinant of the performance of the turbine. As a result, these components are often manufactured in-house to protect IP and guarantee quality (where production is outsourced, WTG producers must ensure that suitable quality is maintained). These components tend to attract the most R&D.

• Input costs. Other components (such as towers) are typically manufactured to a relatively standardized design. In these circumstances it can be easier to outsource the production. For these components, cost and location are the most important factors.

• Transportation. Clearly, some of the components of a turbine present certain challenges in terms of transportation and logistics (typically blades and towers). Investing in local production facilities has been an important objective for WTG producers. Local component sourcing is also an important objective.

• Subcomponent supply. Even where a component is produced “in-house,” this may not shield the WTG producer from supply chain pressures. For example, problems in the supply of bearings for gearbox manufacturing have been well documented. In addition, we believe it is important to distinguish in-house production from in-house assembly. A producer who assembles a gearbox using parts from sub-suppliers may be more exposed than a producer who produces the parts from scratch (i.e., milling their own steel, etc.)

Blades

Blades can now be pitched (turned into the wind to reduce the exposed surface area) to reduce stress load on the tower and control blade speed. Variable speed systems allow turbines to operate at high efficiency over a wider range of wind speed as well as lower stress loads. This not only increases the number of sites that can be exploited, it also reduces costs.

As a high-value-added component with design specifications unique to each specific turbine, these components are often constructed in-house to ensure better control of quality and protection of IP. When outsourced, privately held LM Glasfiber (located in Denmark) is the key industry supplier (more than one-third of installed capacity uses their blades).

With lengths of up to 60m, blades can be difficult and expensive to transport, and the risk of damage in transit requires careful logistics planning. This also requires that turbine manufacturers establish blade production sites near the end market.

Blades were traditionally manufactured using high-strength, low-weight fiber glass. Some newer designs use carbon fiber as an alternative. We understand that the raw materials for blade production are not in abundant supply in India or China. Therefore, blade production in these regions will require imports. Blades represent approximately 21% of the cost of an assembled turbine and are the most labor intensive component of the turbine.

Blade quality can be a sore spot for some in the wind industry. During 2Q08, Suzlon (SUEL IN, INR 113.7, NC) announced that it would recall and refit 1,251 blades on 417 of its 2.1MW turbines due to cracking problems. While Suzlon claimed that blade cracking had only occurred on 45 blades and that the wider recall, which represents nearly the entire installed US fleet, was a precautionary measure, the source of the problem has apparently not yet been determined. Subsequent to this announcement, Edison Mission Energy, a division of Edison International (EIX US, $31.56, Buy, covered by Paul Fremont), exercised the option to cancel the remaining 150 turbines of its 300 turbine order and cited that uncertainty over the source of the blade cracking problem as the determining factor.

Gearboxes

The gearbox is required in order to transfer the low-speed, high-torque power from the rotor blades into high-speed, low-torque power to be fed into the generator to create electricity.

Turbine underperformance as a result of gearboxes failure has been a high-profile issue in the industry. We understand that in a number of the cases where turbines have failed, the failure has been manifested in the gearbox bearings. In these cases, it can be difficult to pinpoint whether the problem lies with the bearings themselves, or whether there is something more fundamentally wrong with the turbine design such that undue stress is placed on the system, and that is simply manifested in the bearings. Turbine manufacturers, gearbox manufacturers and bearing suppliers have been working together to untangle this issue.




As a result, we believe that WTG producers are now more likely to view the gearbox as a key, value-added, component, rather than simply a standardized component. Back in 2005, very few companies produced the gearboxes in-house. Since then, there has been a trend towards vertical integration in this area (e.g., Siemens purchase of Winergy, Suzlon’s purchase of Hansen [HSN LN, 159p, Buy]). However, these “captive” gearbox makers continue to supply third-party WTG producers.

A key issue for gearbox supply in recent years has been a shortage of large bearings. But bearing supply is improving as established bearing manufacturers have begun to address the wind market. Although we expect bearings supply to remain relatively tight in 2008, by 2009 we believe that new capacity should come on line to ease the bottleneck.

In turn, gearbox manufacturers, led by market leaders Winergy and Hansen, are planning a period of significant capacity expansion. In line with the supply bottlenecks and steel price increases that we have seen, prices for gearboxes have been on the rise. Hansen gearbox contracts also have provisions which allow them to pass on increases in other input costs (such as labour costs) in the form of higher prices.

We note that BTM Consult recently announced that they believe that bearing supply remains the primary bottleneck for the wind industry, with the shortage potentially lasting for another two years. It should be noted that they do make a distinction between the severity of shortage for different types of bearings. They suggest that ‘slewing’ bearings, which are used for the turbine’s yaw and pitch mechanisms, will be more tightly constrained than the ‘roller’ bearings which are used for gearboxes.

Gearboxes usually constitute approximately 10%–15% of the total cost of the turbine. Of this, approximately 60% is steel. Compared to towers and blades, the lower weight and size results in less logistic complexity.

Generators, Converters and Transformers

The generator converts the mechanical energy into electrical energy, the converter and transformer convert this electrical energy into alternating, high voltage power required by the grid.

These are relatively standardized components, although turbine manufacturers prefer to source third-party generators from suppliers with previous experience in the wind industry. These components make up approximately 10%–15% of the cost of the turbine. Like generators, a handful of large suppliers dominates the supply market.

Hubs, Shafts and Mainframes

The hub holds the blades in place and is typically made from cast iron. The shaft transfers the energy from the hub to the gearbox. The mainframe sits atop the tower and holds the nacelle in place.

Due to the relatively standardized design, these components are usually outsourced. Supply can be become constrained when demand from other heavy industry sectors increases. We believe that this area will benefit from increased supply from China in the coming years. Given the inputs involved, the key consideration is cost, which normally comprises about 6%–7% of the turbine cost.

Towers

Towers are a relatively standardized product usually made from rolled steel. They can range in size from 40m to over 100m depending on the turbine design.

The tower normally represents about 20%–25% of the total cost of the turbine, and 60%–70% of the total steel consumption. There are two key considerations for WTG producers in this regard. First, rising steel input costs are particularly important. Negotiating steel contracts and developing towers from different materials have been important objectives. Second, given the weight of the towers, transportation and logistics costs are high. Therefore local production is preferred. We understand that this area is unlikely to be supply constrained in the coming years.

Several turbine manufacturers are investing heavily in R&D to reduce tower costs. The principal avenue for tower cost reduction is to reduce tower head (nacelle and blades) mass. Alternatively, other materials such as concrete could be deployed to reduce steel consumption.

Control systems

Software is required to control the angle of the turbine and the pitch of the blades in order to optimise power output in different wind conditions. These systems, which can use complex algorithms designed in conjunction with the dynamics of the blades are developed in house.




Nacelle

The nacelle is normally made from lightweight fiber glass and provides the casing for the turbine drivetrain. This is not a particularly value-added component and normally represents less than 2% of the total turbine cost.

Transport & Logistics

As introduced above, as global sales expand, managing transportation and logistics (inventory) costs will become increasingly important for WTG manufacturers.

EXHIBIT 3: COST DISTRIBUTION

Source: Gamesa

As has been well documented in recent coverage of the sector, the cost of steel is clearly an important component in the total cost of a turbine, particularly in the tower, the shafts and mainframe, and the gearbox. As steel costs continue to rise, this could put pressure on margins if rising prices reach a ceiling (perhaps driven by local incentives). However, whilst we do not disagree with this scenario, we do point out that labour costs are an equally significant component of total cost. Ensuring access to suitable human capital, at the right costs, could be an equally important area of focus in the coming years.

Supply chain constraints

The wind industry has been characterized by supply chain constraints for a number of years. Some of the key points to note are:

− Components suppliers have in the past been reluctant to expand capacity in some areas. There are a number of reasons for this:

• For some components, such as gearboxes and bearings, capacity expansion requires significant levels of investment. The lack of visibility on long-term capacity requirements in the industry, perfectly illustrated by the “on again – off again” nature of the PTC in the US, has made it difficult for suppliers to justify the levels of required investment.

• For items where precise design specifications are required, the risk of technological obsolescence is high. Sub suppliers do not wish to invest in production capacity for a component for a turbine which may be out of favour very quickly. This is less of an issue for the more standardized components.

• Of course, these risks would be more acceptable to the sub-suppliers if they had a chunky margin as compensation. Unfortunately, they do not. Prices and margins in the component businesses have traditionally been relatively low. BTM Consult has suggested in the past that prices increases would be necessary to stimulate capacity investment.

− Another theme is that many of the sub-suppliers tend to be much larger than their turbine producing customers, with the wind industry as a relatively small proportion of their total output. This tends to reduce




the consumer power of the WTG producers as suppliers focus attention on their larger customers. In addition, supply available to the wind sector can also be influenced by the changes in demand of other heavy industry players. This issue has historically been a key factor in the wind industry’s relationship with established bearings manufacturers where total wind demand was at times less than 5% of supplier output. The strong growth of the wind industry is playing a key role in not only giving the wind industry a seat at the head table but also bringing previously uncommitted suppliers such as Timken to the wind industry.

− It is also worth noting that supply issues tend to become more important as turbine size (MW) increases. For example, as the power capacity and weight of the turbine increases, so do the stresses on the components in the turbine. As a result, higher-quality, more reliable components are required. With the industry moving towards larger turbines, this situation should be monitored.

As the industry matures, and particularly as visibility increases, we believe that sub-suppliers will be more willing to invest in capacity expansion.

EXHBIT 4: EXPANSION PLANS FOR SELECTED KEY COMPONENT SUPPLIERS

Component Supplier Expansion plans Bearings SKF − SKF (SKFB SS, SEK 73, NC) recently announced an expansion of large bearing

capacity expansion in its plant in Dalian, China. Capacity at the plant is to be doubled from that detailed when the plant construction was announced in 2005.

Gearboxes Hansen − Expanding production capacity from the current 6,000 MW to 9,000 MW by 2010 and 14,300 MW by 2013.

− Significant proportion of this increase will be in India and China

Blades LM Glasfiber − In 2007, capacity expanded by over 1,000 MW in Spain, China and India. − Currently building new manufacturing facilities in the US and Poland, in addition to

expanding capacities in existing plants in Denmark and China. In total in 2008, capacity should expand by more than 2,000 MW.

− The capacity expansion has not been plain sailing for LM Glasfiber, with ramp up costs and delayed production start-ups contributing to an operating loss in FY2007 (-7% EBIT vs 11% in FY06)


In an oft-cited supply chain study by BTM Consult published in December 2006, it was claimed that there was an insufficient supply of gearboxes and large bearings to meet the level of 2006 demand. At the time, BTM Consult stated, “It is our opinion that 15,100 MW in 2006 will be difficult to reach, and almost 20,000 MW in 2007 is not possible due to supply constraints”. In fact, installations in 2006 and 2007 were 15,016 MW and 19,791 MW, respectively. In our view, this illustrates that the problems in the supply chain can be overestimated.

A similar study has recently been released by BTM Consult to take account of changes to the supply chain, and changes to future demand estimates, since the time of the first report. As Exhibits 5 and 6 show, the situation remains broadly the same. For 2008, it is suggested that bearings will remain insufficient to meet demand, although gearbox supply is less constrained. Looking out to 2012, gearboxes and bearings are predicted to remain two of the most constrained components in the wind supply chain.

EXHIBIT 5: TURBINE COMPONENT SUPPLY 2008

Source: BTM Consult

EXHBIT 6: TURBINE COMPONENT SUPPLY 2012

Source: BTM Consult




Turbine size

The modern turbine has come some way from its 7th century predecessor. In fact, technological development in the last ten years has been significant with a trend towards larger, more powerful, more durable, and more efficient, turbines. As the charts below clearly illustrate, wind turbines have increased dramatically in size and capacity over the past 25 years. This has allowed demand to shift from localized production, often driven by lack of grid access or environmental concerns, to large utility scale developments designed to feed the grid.

EXHIBIT 7: WIND TURBINE EVOLUTION

Source: BTM Consult

EXHBIT 8: AVERAGE TURBINE SIZE (MW)

400

600

800

1,000

1,200

1,400

1,600

2000 2001 2002 2003 2004 2005 2006 2007

Ave

rage

turb

ine

size

inst

alle

din

year

(kW

)

Source: BTM Consult

Despite the failure of the offshore market to really take off, the market for large scale turbines has shown strong growth. For example, the share of >2.5MW turbines in annual installations has more than doubled from 2.4% in 2005 to 5.3% in 2007.

EXHIBIT 9: 2007 WTG INSTALLATIONS BY MW CLASS

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

"Small WTGs" < 750 kW "One-MW" 750 - 1500 kW "Mainstream" 1500 - 2500 kW "Multi-MW class" > 2500 kW

Shar

eof

glob

alin

stal

latio

ns20

07

Source: Jefferies International, BTM Consult

From the producers’ perspective, larger turbines result in scale economies in terms of cost:

• The cost of some components is independent of the size of the turbine (for example, the control system).

• Other costs do rise on a linear basis with the increase in the size of the turbine. We believe that this is the case for the key raw material steel although the industry is looking to reduce nacelle mass as well as seeking solutions to reduce steel usage in the tower (the tower represents approximately 60%–70% of total steel consumption of a wind turbine)

As a result, the cost per MW of production reduces as turbine size increases. Assuming a simple “cost-plus” approach to pricing, this would suggest that the price per MW would fall as turbine size increases. However, we believe that in fact the reverse is true — as turbine size increases, the price per MW also increases. We suggest the following explanations for this:




Ø = Rotor Diameter

50 kWØ 15m

100 kWØ 20m

500 kWØ 40m

600 kWØ 50m

2,000 kWØ 80m

5,000 kWØ 124m

1980 1985 1990 1995 2000 2003

Ø = Rotor Diameter

50 kWØ 15m

100 kWØ 20m

500 kWØ 40m

600 kWØ 50m

2,000 kWØ 80m

5,000 kWØ 124m

1980 1985 1990 1995 2000 2003

• Economies of scale on the purchasers’ side also exist. For example, creating a 100 MW wind farm using 50 two-MW turbines instead of 100 one-MW turbines requires significantly lower installation and maintenance costs. Therefore a higher price can be paid per MW for a larger turbine.

• We believe that the pricing of turbines may be more influenced by the annual power output and hence revenue generating capacity as measured by mega-watt hours, rather than by the power capacity of the turbine measured by mega-watts. Depending on the characteristics of the wind site, the MW/h annual capacity of a 2MW turbine can be more than double that of a 1MW turbine. This reflects that larger, more advanced turbines capture more wind energy and thus generate more output per capacity than smaller, older turbines.

Offshore Wind

As of YE2007, there were 1,077MW of operating offshore wind farms located primarily in the UK and off the coast of Denmark. Most of these operate in relatively shallow water where installation and operational costs are easier to control.

The offshore market should offer great potential for growth as much of Europe’s best wind resources are located off the coasts and therefore Not in My Backyard (NIMBY) concerns are expected to be far lower if the turbines are placed several miles off the coast and mostly out of sight.

As the chart shows, the offshore wind market has failed to build on the early promise shown in 2002/03 when the first offshore farms were installed. Since that time annual offshore installations have totaled only about 1% of total installations. BTM expect the market to become more significant in 2009 and beyond.

EXHIBIT 10: OFFSHORE INSTALLATIONS

0

500

1,000

1,500

2,000

2,500

2002 2003 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2012E

Ann

ualo

ffsho

rein

stal

latio

ns(M

W)

0%

1%

2%

3%

4%

5%

Shar

eof

tota

lins

talla

tions

Annual offshore installations (MW)

Share of total installations

Source: BTM Consult

There are a number of factors holding back the growth of this market.

− First, there are significant technical challenges related to the offshore market. The cost of installing an offshore turbine is estimated to be 50% higher than onshore and operating / maintenance expenses are 100% higher.

− In order to make the economics work, offshore installations must receive privileged incentive programs or operate at potentially unrealistically high capacity factor and/or turbines must be offer significantly higher power output. Turbine manufacturers are working to develop larger, “second generation” (5-7.5 MW) turbines which could deliver the economics for offshore to work. However, as the turbines get larger, the stresses are greater.

− We understand that financing and approval is more difficult for offshore sites.

We believe the challenge will be to develop deep water sites and that this market is unlikely to show significant growth for another three to four years.




Wind Market Background

Emergence of wind power

The wind market emerged in the 1980s in both the U.S. and Europe in response to the oil crisis of the 1970s, which led many countries to seek alternative sources of energy. However, it was not until the middle of the following decade that the modern wind industry began to emerge. As is shown in the chart below, the global wind market has gone from strength to strength with cumulative installed capacity increasing at a 29% CAGR over the past 10 years. Wind now represents the largest share of alternative energy generation and represents up to 10% of electricity consumption in wind-intensive markets such as Germany.

EXHIBIT 11: CUMULATIVE GLOBAL MW INSTALLED – HISTORIC (1997–2007) AND FORECAST (2008E–2012E)

0

50,000

100,000

150,000

200,000

250,000

300,000

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2012E

Cum

ulat

ive

MW

inst

alle

dat

year

end

Asia

Americas

Europe

10 year historic CAGR 29%

5 year forecast CAGR 25%

Source: BTM Consult.

We believe that the future looks very bright for the wind industry, as reflected in the BTM forecasts shown in the chart above, which predict a 25% CAGR over the coming five-year period. Its relatively low cost and widespread availability make it an ideal solution to the need to reduce dependence on energy imports and lower harmful emissions from the energy generation industry. In addition, the rising cost of energy and the potential for carbon emission charges could narrow the gap between traditional fossil fuel generation and wind generation. Indeed, in certain areas wind can already compete with traditional electricity on a level playing field. However, in most locations, wind will require incentives in order to compete with traditional electricity. We believe these incentives, in whatever form they appear, likely will continue for the following reasons:

� Competition for resources. The increased prices for fossil fuels have been in part driven by increased competition from the developing economies as well as rising demand in the developed economies. While conservation could offset higher demand in the developed world, we do not expect demand from the emerging economies to slacken and thus prices for traditional fuels will continue to face pressure.

� Energy security. Many western nations rely on imported fossil fuels. While wind power cannot reduce oil dependency, it can help reduce demand for the natural gas used in the growing number of gas fired generators. Unfortunately for Europe and the U.S., many fossil fuels are found in regions that are unfriendly, unstable, or unreliable, which forces governments to seek home-grown renewable sources of energy.

� Environmental concerns. Awareness of man’s impact on the environment appears to be on the rise, and global warming in particular is attracting plenty of media attention as of late. Wind has one of the lowest emissions of any alternative energy and can be deployed in massive scale.

� Falling cost premium. The sustained rise in traditional energy prices (4%–5% p.a. over the last 10 years in the US and Europe) has reduced the premium paid for alternative energy and weakened the key argument against the deployment of alternative energy resources. Furthermore, if the environmental impact of fossil fuel pollution was added to the cost of traditional energy generation, the alternative energy premium would shrink significantly and even disappear in many places. Oil prices in excess of $100/barrel and its impact on natural gas prices, as we have seen in 1H08, only serve to reduce this premium further.




EXHIBIT 12: WIND POWER SHARE OF GLOBAL ELECTRICITY GENERATION

0%

1%

2%

3%

4%

5%

6%

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Actual Forecast

Source: BTM Consult, IAE World Energy Outlook 2007

Incentive Programs

As previously mentioned, incentive programs rather than wind characteristics determine the growth (and the prices) of wind installations, particularly in Europe and the United States. Although, once installed, a wind farm has relatively low operational costs because there is no requirement for fuel, like virtually all forms of alternative energy, wind power suffers from intermittency issues. In other words, the blades only turn when the wind is blowing and utilization (“capacity factor”) is much lower than a traditional gas or coal burning plant. Thus a wind farm generally requires some form of incentive in order to generate an attractive financial return, although rising energy prices mean that wind farms in advantageous locations can compete without any form of subsidy.

Incentive programs vary from country to country and, within the US, from state to state. Given that two of the largest wind markets in the world, Spain and Germany, both utilize a form of feed-in tariff, it would be logical to assume that this is the most effective catalyst for wind investment. We will discuss three primary forms of incentives:

� Feed-in Tariffs. These tariffs guarantee that 100% of produced power will be purchased by the grid operator at a fixed or pre-agreed price. The structure of the tariff varies from country to country and can be set either as a fixed rate or as a premium over a pool price. Another key factor is the duration of the feed in tariff agreement as the longer the wind farm operator can secure a guaranteed (and hopefully premium) price on the output, the better the return.

� Renewable Portfolio Standards (RPS). Federal or state governments can set hurdle rates for renewable generation as a percent of the total and levy fines for non-compliance. This has led to tradable Renewable Obligation Certificates (ROCs) in the United Kingdom which allow producers of low emission energy to receive a premium on their energy sales paid for by fossil fuel generators who have not met their renewable obligation.

� Production Tax Credit (PTC): The U.S. federal government will offer a $0.018 tax credit for every kWh of energy produced in the first 10 years of operation. In October 2008, following months of political wrangling, the PTC was renewed for a further 12 month period until December 2009.

Additional details on the specific incentive arrangements in a number of countries are provided in the Appendix.




Europe leads the way …

While the modern wind industry has its roots on both sides of the Atlantic, as for many other sources of alternative energy, it was Europe that quickly emerged as the global leader in wind powered generation. Europe represented approximately 60% of all wind installations in the 10-year period to 2007. This historical trend is not a result of a windier Continent. Indeed, the U.K. has some of the best wind potential in Europe while the American Midwest has been referred to as the “Saudi Arabia of Wind.” As we have discussed, the difference is the incentive programs that allow investors to earn an attractive return on wind projects.

EXHIBIT 13: EUROPEAN ANNUAL INSTALLATIONS – HISTORIC (1997 – 2007) AND FORECAST (2008E – 2012E)

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2012E

Ann

ualM

Win

stal

latio

ns



Source: Jefferies International, BTM Consult.

While Europe currently represents 70% of global installed capacity, Germany and Spain alone account for 65% of that capacity (or 40% of total global capacity). Most countries have relatively minor investment in wind despite their often times superior wind characteristics. The chart below illustrates the average wind speeds throughout Europe. In particular, we suggest that investors focus on the British Isles and northern France where some of the best wind conditions are located and, ideally, are near centres of demand.

− France has an incentive scheme which allows for a 10-year feed in tariff at €0.082/kWh and has a target of 13,500MW of wind capacity by 2010 from a current 2,471MW at YE07.

− As the chart below illustrates, the UK has some of the best wind resources in Europe and, given its high population density, is likely to be a leading catalyst for development of offshore wind. The Crown Estate, the public body which controls the seabed around the UK, has recently launched round three of its leasing program for the delivery of up to 25 GW of offshore wind farm sites by 2020. This builds on the 8 GW under development from rounds one and two. Although not intending to own or operate any of the farms, the Crown Estate is planning to co-invest up to half of the cost for permitting, supply chain constraints and grid connection, in order to speed up wind farm delivery.

− In Germany, as part of the recent amendments to the EEG legislation which confirmed the reductions to the solar feed-in tariffs, the feed-in tariffs for wind energy were increased, to take effect as of 1 January 2009. The tariff for onshore projects was increased from €0.08/kWh to €0.092/kWh, with a subsequent 1% annual reduction for new projects. Offshore projects, for which Germany has a healthy wind resource, will receive a €0.15/kWh tariff, which is guaranteed for projects started in the period to 2015. The legislation also contains special premiums for repowering projects and stricter obligations for grid operators aimed to facilitate the integration of more wind energy into the grid.




EXHIBIT 14: AVERAGE EUROPEAN WIND SPEEDS

Source: Risø National Laboratory, Denmark

… while the Americas are catching up

While the average growth rate of cumulative installed capacity is relatively strong, the lack of consistency has been a significant limiting factor. The volatile nature of annual installments is linked to the on-again off-again nature of the PTC. A reading of the charts below clearly shows the impact on US installations of a failure to have the PTC in place and the impact that this inconsistency has had on Vestas (VWS DC, DKK 356, Buy), which has been a leading supplier into the US market over this period.

EXHIBIT 15: AMERICAS ANNUAL INSTALLATIONS

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2012E

Ann

ualM

Win

stal

latio

ns




EXHIBIT 16: PTC EFFECT ON VESTAS’ MARGINS

-2,000

-1,000

0

1,000

2,000

3,000

4,000

5,000

6,000

2001 2002 2003 2004 2005 2006 2007

Ann

ualA

mer

icas

inst

alla

tions

(MW

)

-4%

-2%

0%

2%

4%

6%

8%

10%

12%

Vest

asEB

ITm

argi

n

American installations (MW) - PTC in place

American installations (MW) - PTC NOT in place

Vestas EBIT margin

Quality issues lead to falling margins in 2005


The good news is that 2007 was a record year for installations in the US and we believe, along with BTM Consult, that this trend is likely to continue. With developers looking to finalise installations before the original expiry date for the PTC in December 2008, we believe 2008 will be another bumper year. Likewise, with the PTC now extended for an additional 12 months, 2009 should also be a very strong year for US demand.

The US market is particularly attractive as it not only has excellent wind resources (the American Midwest has been referred to as the “Saudi Arabia of Wind”) but plenty of domestic demand and a well established grid. In addition, many potential wind farm sites are located in areas of low population density and, phrasing this politely, minimal chance of disruption to sites of outstanding natural beauty (your analyst lived in the Midwest for 12 years and can attest to this personally).




>7,5 M/seconde6,5 à 7,5 M/seconde5,5 à 6,5 M/seconde4,5 à 5,5 M/seconde

<4,5 M/seconde

>7,5 M/seconde6,5 à 7,5 M/seconde5,5 à 6,5 M/seconde4,5 à 5,5 M/seconde

<4,5 M/seconde

US installations are still concentrated in a handful of states such as Texas and California. However, the spread of state RPS should lead to an ever increasing pool of ‘pro-wind’ states.

EXHIBIT 17: US WIND RESOURCE

Source: NREL

EXHIBIT 18: US INSTALLATIONS BY STATE

Source: NREL

Wind represents an increasingly important source of new generation capacity in the US as 2007 was the third consecutive year that wind generation capacity was the second largest generation resource added to the US grid following natural gas and well ahead of third place coal.

EXHIBIT 19: SOURCES OF NEW US GENERATION CAPACITY

Source: EIA, Ventyx, AWEA, IREC, Berkeley Lab

The future for wind looks very bright, as is illustrated by NREL data gleaned from the interconnection queue from eleven of the most wind-significant independent system operators and regional transmission authorities. Total wind projects in the transmission queue account for 225GW of new capacity, or more than 12 times the current installed wind capacity. Furthermore, wind is by far the largest source of power in the interconnection queue and is more than twice the total of the second largest source, natural gas. As a reminder, while entering the interconnection queue is a pre-requisite to bringing a project on-line, not all projects in the queue will reach fruition. Additionally, while the ISO and RTAs selected for this survey represent only 60% of new annual US capacity installations, they represent a much higher share of wind installations.




EXHIBIT 20: ANNUAL INSTALLATIONS AS % OF TOTAL US INSTALLATIONS

Source: NREL, Exeter Associates

Asia arrives

India and China have emerged as growing forces in the global wind market, making up nearly 15% of installed global capacity as at YE2007. While much of the developed world views wind as a clean source of energy, these countries suffer from power deficits and use wind both to reduce harmful emissions as well as to grow their installed base to meet rising demand.

EXHIBIT 21: ASIA ANNUAL INSTALLATIONS – HISTORIC (1997 – 2007) AND FORECAST (2008E – 2012E)

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008E 2009E 2010E 2011E 2012E

Ann

ualM

Win

stal

latio

ns




The forecast growth rate of 25% per annum compares favorably to the forecasts for the more mature European and US markets (18% and 20%, respectively) discussed previously. To some extent, this forecast growth rate is distorted by a low base period. As the chart below shows, BTM expect that the growth will be front loaded and that from 2009E onwards the growth rate in Asia will decline to the 15%–20% global industry average.




EXHIBIT 22: MW INSTALLATIONS BY REGION – ANNUAL GROWTH RATES

0%

10%

20%

30%

40%

50%

60%

70%

2006E 2007E 2008E 2009E 2010E 2011E 2012E

Americas

Europe

Asia


Looking at China in particular, the chart below illustrates how the rise of domestic Chinese WTG producers has reduced the market shares of both Gamesa (GAM SM, €18.13, Buy) and Vestas from above 35% in 2004 to around 15% in 2007.

EXHIBIT 23: SHARE OF ANNUAL CHINESE INSTALLATIONS

0%

5%

10%

15%

20%

25%

30%

35%

40%

2004 2005 2006 2007

Other China

Goldwind

Gamesa

Vestas





National Wind Markets

USA

Historically, installations in the US have fluctuated with the existence of the crucial PTC wind incentive. I2007 saw record installations as the PTC remained in place. The strong level of demand is expected to continue in 2008 as wind developers seek to finalise projects before the year end, when the PTC was originally scheduled to expire.

On October 3rd, HR1424 (the financial "bailout" package) was signed into law with an attached bill that includes a twelve month extension of the PTC. This should secure the US as one of the leading wind markets into 2009.

We believe that a longer term wind incentive (perhaps a multi-year extension of the PTC) may be part of the incoming Administration’s energy policy in 2009.

EXHIBIT 24: U.S. INSTALLATIONS 2001–07 (MW)

0

1,000

2,000

3,000

4,000

5,000

6,000

2001 2002 2003 2004 2005 2006 2007

Source: BTM Consult

Germany

Germany returned to a period of declining installation in 2007 after a surprising bump in 2006. Given that the incentive programs remained unchanged and that re-powering (replacing older turbines — usually located in excellent sites — with newer models) and offshore deployment has not yet begun in significant volume, the weaker result did not come as a surprise.

Germany is the first country to break the 20GW of installed wind capacity barrier, and wind power represented approximately 7.6% of total electricity consumption in 2006. BTM Consult expects that German installations will revert to their downward course in the next two to three years before re-bounding, driven by offshore development and re-powering.

EXHIBIT 25: GERMAN INSTALLATIONS 2001–07 (MW)

0

500

1,000

1,500

2,000

2,500

3,000

3,500

2001 2002 2003 2004 2005 2006 2007

Source: BTM Consult

Spain

The Iberian giant (at least in terms of installed capacity) surged in 2007 after the legislative gridlock over feed in tariffs was resolved in 2006. Furthermore, Spain has been moving to strengthen its transmission network to enable further exploitation of attractive wind resources in windy Galicia, and should be on target to reach its target of 20 GW of cumulative installed capacity by 2010E. Following the near doubling of annual installations in 2007, wind energy contributed 9% of total electricity consumption in 2007, which represents a larger share than hydropower.

EXHIBIT 26: SPANISH INSTALLATIONS 2001–07 (MW)

0

500

1,000

1,500

2,000

2,500

3,000

3,500

2001 2002 2003 2004 2005 2006 2007

Source: BTM Consult




France

France emerged as a potentially significant source of new demand growth in 2006 when the French government reversed course and substantially improved the financial terms available to wind farm operators. Traditionally, the French government had argued that its heavy use (more than 3/4 of total generation) of emissions-free nuclear power excused it from the need to support renewable energy. However, the government changed its mind and has targeted 13,500 MW of wind power by 2010, although this target is unlikely to be reached. Installation growth was subdued in 2007 as developers adapted to the new regulations that came in to effect last year, and the medium term looks encouraging for French wind.

EXHIBIT 27: FRENCH INSTALLATIONS 2004–07 (MW)

0100200300400500600700800900

1,000

2004 2005 2006 2007

Source: BTM Consult

United Kingdom

The U.K. installation slumped in 2007, although several large-scale projects were given planning permission, which has lifted the national consented portfolio to nearly 5000 MW. While the growth rate is impressive, the level of installation is low, particularly when one considers the excellent wind resources available in the British Isles, particularly in the north of Scotland.

There were some policy wobbles that have raised concerns over the medium-term outlook for the British wind industry. The Department of Trade and Industry is worried that the current Renewable Obligation (RO) scheme is not delivering sufficient investment. Several policy trial balloons have been raised with varying results.

EXHIBIT 28: UK INSTALLATIONS 2001–07 (MW)

0

100

200

300

400

500

600

700

2001 2002 2003 2004 2005 2006 2007

Source: BTM Consult

India

Indian installations slowed a bit in 2007 due to rising interest rates and higher turbine prices. Demand in India is driven not only by federal renewable energy quotas, but also the strong demand for additional power stemming from the country’s high growth rate and low electrification levels. India is an unusual market in one key aspect. Generally speaking, wind farms are developed to feed the grid. In India, roughly 70% of the wind capacity is owned by manufacturing firms looking to secure their own source of power. The future of wind power looks quite promising, although lack of transmission capacity and failure to deliver sufficient incentives at the state level could limit the potential of the Indian market.

EXHIBIT 29: INDIAN INSTALLATIONS 2001–07 (MW)

0200400600800

1,0001,2001,4001,6001,8002,000

2001 2002 2003 2004 2005 2006 2007

Source: BTM Consult




China

China’s emergence has been even more spectacular than India although cumulative installations have not yet reached the level on the sub-continent. This growth has been driven by the central government’s target of 10% of electricity from renewable resources by 2010 and 15% by 2020. Additionally, China has a voracious appetite for additional power so, much like India, market forces are very much at play along with regulatory incentives. We would expect that China will likely continue to be a significant source of demand. China, like many emerging wind markets, has mandated a minimum local content requirement (70% in China) which has resulted in the emergence of several domestic turbine manufacturers led by Goldwind, although foreign producers still control the majority of cumulative market share.

EXHIBIT 30: CHINESE INSTALLATIONS 2001–07 (MW)

0

500

1,000

1,500

2,000

2,500

3,000

3,500

2001 2002 2003 2004 2005 2006 2007

Source: BTM Consult

A source of concern in China is the potential quality issue of the installed turbine fleet and wind farms. According to the China Wind Energy Association (CWEA), nearly one-third of all installed turbines are not connected to the transmission grid. While we should assume that some portion of these turbines are captive capacity designed for immediate local use, it is clear that China faces a risk of an overheated market that could result in stranded assets if grid connections are not extended. On that front, it is concerning that the most recent energy policy does not mention a specific renewable energy mandate (RPS) for grid companies and could reduce their incentives to invest in the capacity to bring these installed wind farms online. We understand that this issue is widespread and well recognized at the national level, but the translation of national policy into local action is not always a smooth issue.

Another development is that domestic manufacturers now represent a majority of new installations in 2007. Led by Goldwind (002202 CH, CNY 17.5, NC) and Sinovel, domestic producers captured 56% of the new turbine sales in 2007 although foreign and JVs still represent the majority of cumulative installed capacity. A part of this shift is attributed to the 70% local content rules for federal projects. Also, average turbine size in China is much smaller than Europe or the USA, while many of the established Western turbine producers are focused on MW+ scale turbines.




AppendixEXHIBIT 31: INCENTIVE PROGRAMS

Country Rate per

kWh Euro-cent

Comments

Austria 7.65 The purchase price for wind power in 2007 was €0.0765/kWh, which was €0.0755/kWh in 2006. This tariff is guaranteed for 10 years.

Australia 4.45 The Renewable Energy Act in force since 2001 regulates the subsidies used to promote wind energy. Price at EURct 4.4 per kWh divided almost half & half between energy price and environmental bonus.

Belgium 7.50 The tariff consists of 5.0 EURct (fixed for 10 years) + 2.5 EURct (green certificate).

Czech rep. 8.13 – 10.16

Two options: a fixed price for electricity or a green power bonus on top of the wholesale electricity price. Both options last 15 years. The first option is popular in Czech. Under the first option, project that came into operation in 2007 get €0.1/kWh this year. Those came into operation in 2008 get €0.09/kWh. The green bonus for wind projects came into operation in 2007 is €0.08/kWh, for those in 2008 is €0.075/kWh.

Denmark 6.93 In proposal now. New installations get Nordpool price plus DKK 0.25 per kWh in 22.000 full load hours thereafter only Nordpool price. This figure is subject to approval by the EU.

Estonia 7.35 – 8

There are 2 options: 1. To sell the electricity under 12 years contract on either market price of 2.6 EURct/kWh plus a premium of 5.4 EURct/kWh. OR 2. To sell the electricity under 12 years contract to Eesti Energia with a feed in price of 7.35 EURct/kWh.

France 8.20

Under the current regulation, the government’s purchase prices for wind power are: Onshore plant is €0.082/kWh for the first 10 years, after which it varies between €0.028kWh for plant operating at full capacity for an average of at least 3600 hours a year and up to €0.082/kWh for 2400 hours or less of operation.

Germany 8.8 – 9.1

8.8 EURct per kWh onshore & 5.9 EURct per kWh onshore. Feed-in law as of July 2004. The rates will apply for the initial 5 years of operation and thereafter the overall feed-in rate will be adjusted to reference “energy-values” for the respective location. Prices initial 5 years at 8.36 EURct thereafter 5.28 EURct per kWh. Poor wind site will prolong the higher rate. Off Shore projects have for the initial 9 years EURct 9.10 thereafter EURct 6.19 per kWh.

Greece 6.61 – 8.17 Rates are different by location. Mainland wind energy producers get 90% of the consumer price. Wind farms without grid access to mainland get 7.31 cents per kWh. A 40% grant of capital costs possible. Special rates available on the Greek Islands.

Ireland 5.7 – 5.8 Projects larger than 5 MW get 5.7 EURct and smaller installations get 5.8 EURct. Tariffs on Ireland are index regulated and there is a 15 year contract period.

Italy 18.50

6 EURct per kWh Electricity price. 12.5 EURct per kWh Green Certificate. Tariff for 2006 production was 14.94 EURct/kWh. Recent plants now come under the new support scheme based on a compulsory RES quota and tradable green certificates granted to qualified RES plants over 12 years of lifetime. Certificates relating to the 2006 production are trading at appr. 12.5 EURct/kWh, which adds to a wholesale price of electricity of about 6 EURct/kWh.

Japan 8.86 Rates for projects realized in 2002 and 2003. Further subsidies up to 50% of the capital cost can be obtained for public companies and 33% for private companies.

Latvia 2.54 Tariff appr. EURct 2.5 as per pool price.

Lithuania 6.30 The guaranteed purchase price for wind power is LTL0.22/kWh (€0.064/kWh) valid until 2020. In Feb 2008, the government was to rule on a proposal to raise the price to €0.087/kWh.

Netherlands 8.00 – 9.00

After the stop of the MEP in 2006, a replaced program Stimuleringsregling Duurzame Energie (SDE) will come into force from April 2008 as part of a new national energy strategy. SDE will guarantee a purchase price for wind power, €0.088/kWh at 2008, which includes a subsidy €0.028/kWh for onshore paid for the first 15 years of operation. For offshore wind, subsidy rate will be granted in 2009.

Norway 5.17 EURct 4.0 kWh Nordpool price app. plus EURct 1.0 kWh as subsidy. It is considered that the investors will get a 15-year support. New feed in price is under consideration. Enova is the administrating body.

Poland 9.97 EURct 9.0 pr kWh. The price consists of energy and a subsidy for a green Certificate.

Portugal 8.00 Price for wind power remain varied from €0.069 to €0.093. The average wind power purchase prices are around €0.08.kWh.




Country Rate per

kWh Euro-cent

Comments

Spain 9.34

EURct 9.34 per kWh variable. EURct 6.8929 per kWh as fixed. Rates are obtained adding a premium of 3.0635 EURct/kWh plus an incentive of 0.7659 EURct/kWh to the market price with a lump sum that averaged 9.34 EURct/kWh during 2006. Generators have also the option to get a full fixed price of 6.8929 EURct/kWh for their electricity. Spanish legal framework is currently under revision to reduce such high prices. New regulation expected within the first quarter of 2007.

Sweden 7.55

Onshore: 7.6 EURct & Offshore: 88 EURct. The compensation consists of the following main components: Electricity payment per kWh 4.9 EURct + “Elcertificate” 2.1 Ect + Onshore environmental bonus 0.4 EURct (or Offshore environmental bonus 1.6 EURct) + Grid benefit reimbursement 0.2 EURct.

UK 6.58 – 7.01 Quota system for all RES-E. Starting at 3% in 2003 up to 10.4% in 2010 – there is a penalty for non-compliance at 3.51 £/kWh. Eligible RE S-E are exempted from the Climate Change Levy certified by the Levy Exemption Certificates. These cannot be trade separated from the electricity. The levy tax is 0.43 £/kWh.

US 3.11 PTC is in force until end 2008. Renewable Portfolie Standards now in 21 Federal States.





EventThis report provides a broad overview of ethanol and other biofuels,both the pitfalls and potential opportunities. We highlight several keyfactors that are likely to drive the biofuels industry over the next threeto five years.

Key PointsTwo approaches to a 16bn gallon/year industry. For the nearterm, we believe the sector will trade off news flow, particularly oil &gas prices (positive if rising), feedstocks (negative if rising), andregulatory news (positive for second-generation fuels). Forlonger-term investors, we believe the underlyingfundamentals—competitive advantage, supply/demand of feedstocksand related costs, as well as cost structures—should drive success orfailure.

• Focus shifting to cellulosic ethanol: With ethanol on track toconsume almost a third of U.S. corn production in 2008-2009, andsupport for corn ethanol mandates stalling, we expect investors tofocus on companies that can decouple biofuels from agriculture andsidestep "food vs. fuel" risk. Tailored energy crops, cellulosicethanol production using non-food crops, or integrated biorefinerymodels could receive premium valuations, and have a betterchance at long-term success.

• Consolidation in U.S. corn ethanol. Depressed margins, risingborrowing costs and shifting political winds argue for a wave ofconsolidation while the industry waits for the distributioninfrastructure to catch up. With planned capacity additions likely tolift capacity to 13.4m gal/year, outrunning the RFS mandate until2012, we expect sector margins to remain under pressure. Likelywinners combine economies of scale and cost advantages insourcing, production, and transportation. Enabling technologiesshould also do well, particularly seed and enzyme companies.Companies that offer retrofit solutions could see more volatility dueto customer credit risks.

• New technologies complicate picture. A wider range ofinvestment themes are gaining traction, including biocrude,biobutanol, cellulosic ethanol, and biomass-to-liquids. Technologyvendors could be whipsawed by commodity market dynamics,including feedstock volatility, rising capital costs and more stringentlending standards.

October 6, 2008

Clean TechnologyIndustrial Biotech

Clean TechnologyBiofuels Primer

Investment SummaryBiofuels continue to attract investor interest, as new technologies,feedstocks and applications provide ways to create value whileside-stepping the "food vs. fuel" debate. Short-term issues that couldimpact stocks, such as difficulty accessing capital markets andvolatile feedstock costs, need to be balanced against longer-termfundamentals, especially government mandates.


Robin Campbell, Ph.D., Equity Analyst44 (0) 20 7029 8678, [email protected]

Lucy Watson(212) 284-2290, [email protected]


Overview

Support for corn ethanol at risk. The biofuels landscape continues to be overcast by thematic concerns. If the first half of 2008 was dominated by the “food versus fuel” controversy we highlighted in 2007, the hallmark of the second half of the year has been, and will probably continue to be, project delays and difficulty accessing the capital markets (most notably Verasun suspended its equity offering and indicated it is considering offers from multiple parties). Election year politics, which had once appeared to be a “guaranteed” catalyst for the sector as candidates appeal to Iowa, has turned into a headwind, with Senator McCain explicitly proposing the elimination of U.S. ethanol subsidies. The perception of heightened political risk could affect the discount rates applied to biofuels until after the next election, at least.

EXHIBIT 1: SUMMARY OF KEY THEMES IN BIOFUELS ($/GAL)

Year Theme Valuations on expected capacity M&A Capital Costs

2006Scarcity:

MTBE & Mandates Drive Expansion $3-$4 $1.74-$3.85 $1.50-$1.75 2007 Consolidation Begins $1.15-$1.75 $1.65-$2.56 $2.00-$2.25 1H08 Food vs. Fuel & Project Delays $1.50-$1.75 (given delays) $1.60 (USBE) $1.75-$2.25 2H08 Credit Crunch $0.70-$1.00 ? $1.75-$2.25 2009 Alternate Feedstocks & Co-Products 60%-80% of replacement value ? ?


On to the next new thing. Overall, however, we believe that the roadmap we sketched out in prior reports continues to apply. In particular, as more industrial biotechnology companies come public and move toward profitability, we expect ethanol shares to stabilize around 60%–80% of replacement value. As a result, investors are likely to focus on next-generation technologies. In particular, we would highlight the following themes:

• Cellulosic ethanol. Cost estimates have been coming down quickly, and we expect the first commercial projects to be announced in 2009–2010. We expect the initial focus to be on sugarcane and energy cane, with switchgrass, algae and, at the other end of the evolutionary spectrum, trees, as potential feedstocks that should be able to overcome the logistical constraints that hamper the sector.

• Biodiesel from waste feedstocks. Lower-cost facilities should be able to sustain a profitable arbitrage against companies that process soybean oil or palm oil. Biodiesel has two significant advantages: it can be used in the existing petrochemical infrastructure, whereas ethanol needs to be blended downstream (effectively requiring the construction of a parallel distribution network), and projected capacity additions for the most part involve feedstock-flexible technology, reducing vulnerability to shortages in a particular feedstock.

• Regional biofuels. Over time, comparative advantage should favor biofuels production in areas with ample water supplies and favorable soil quality, particularly Brazil, Argentina, Indonesia, and Africa. In particular, we believe Brazilian sugarcane-based biorefineries that integrate sugar-based and cellulosic ethanol, and possibly biodiesel and biochemicals, should dominate the lower end of the cost curve.

• Alternate fuels. The 2007 Energy Act has specific mandates for domestic supplies of corn ethanol (peaking at 15m gpy in 2015, vs. current installed and planned capacity of 13.4bn gpy), cellulosic ethanol (rising to 16bn gpy by 2022), and biodiesel (up to at least 1bn gpy by 2012). There is also a mandate for advanced biofuels (rising to 21bn gpy in 2022 from 0.6bn gpy in 2009), a more flexible category which encompasses any biofuel except corn ethanol that reduces GHG emissions by 50%: cellulosic ethanol, biodiesel, biogas, butanol, and other biomass-based fuels (e.g., hydrocarbons from cellulose). The mandate, along with $500m in DOE grants for 2008–2015, should spur a flurry of competing claims concerning the relative viability of different biofuels and feedstocks.

• Traits & capabilities. The value of seed traits has risen sharply over the past few years as Monsanto (MON, Buy, $156 PT) and its competitors have demonstrated that they can extract value by bundling performance-enhancing traits with elite germplasm. We expect the market in 2008–2009 to differentiate between companies that address only one or two steps in the value chain and companies that have both the traits and the capability to insert them effectively into the targeted energy crops. Integrated platforms, in our view, remain the key to capturing a reasonable share (25%–30%) of the value created for farmers/final consumers.

Industrial Biotech

Laurence Alexander, CFA, [email protected], (212) 284-2553


Page 70 of 212

EXHIBIT 2: CORN YIELD TRENDS (BUSHEL/ACRE)

1990 2000 2005 World Average 59 70 75 USA 113 137 149 Argentina 60 93 109 China 74 78 80Brazil 33 47 54 India 23 29 31Sub-Saharan Africa 22 24 25

Source: Monsanto, Doane Forecast

EXHIBIT 3: OIL AND BIOFUEL FEEDSTOCK PRICES, INDEXED YE04=100. THE GLOBAL INVESTMENT IN BIOFUELS WAS ONE FACTOR CONTRIBUTING TO A RISE, AND FALL, IN THE REQUISITE FEEDSTOCKS

Source: Renewable Fuels Association

EXHIBIT 4: FEEDSTOCKS HAVE FALLEN IN SYNCH WITH CRUDE OIL

CAGR % from 2003-2008Commodity Current 1998-2008 2003-2008 Peak TroughEthanol ($/gal) $2.50 7.66% 16.28% -35% 126%WTI ($/bbl) $101 17.36% 31.29% -28% 129%Corn ($/bushel) $4.84 -0.34% 13.56% -31% 157%Palm Oil ($/t) $878 3.237% 17.174% -34% 150%Soy ($/bushel) $9.80 2.19% 9.45% -39% 92%Tallow ($/lb) $0.34 5.72% 14.53% -29% 147%

Source: Bloomberg

Industrial Biotech


0

50

100

150

200

250

300

350

Dec

-04

Feb-

05

Apr-

05

Jun-

05

Aug-

05

Oct

-05

Dec

-05

Feb-

06

Apr-

06

Jun-

06

Aug-

06

Oct

-06

Dec

-06

Feb-

07

Apr-

07

Jun-

07

Aug-

07

Oct

-07

Dec

-07

Feb-

08

Apr-

08

Jun-

08

Aug-

08

WTI CRUDE CORN SOY TALLOW PALM OIL


Page 71 of 212

Key Context for Generalists

Biofuels such as ethanol, biodiesel, and eventually biobutanol are liquid fuels for transportation made from biomass. In effect, biofuels provide a way to convert into a useful form the chemical energy stored in algae, animal fats, or crops such as corn, wheat, beets, vegetable oils (soy, palm, rapeseed), wood, or straw. Biodiesel can be used as a direct substitute for fossil diesel, either as a blend or on its own, while ethanol is used both as an oxygenate (helping gasoline burn more efficiently) and, particularly in Brazil, as a substitute for gasoline. Whereas Brazil has been able to leverage low-cost agricultural feedstocks and relatively low energy use/capita to make a significant shift in favor of using ethanol as a fuel, the industry has been slower to expand in the United States due to lower gasoline prices and the use of other fuel additives (most notably MTBE, which displaced roughly 2bn gal/year of ethanol demand until it was phased out in 1H06). Ethanol has more high-profile support in the Americas. It is derived from sugar, which in turn can be generated from corn (in the U.S.), sugarcane (in Brazil), wine (Sweden), wheat, beets, or even cellulosic waste products such as stems and stalks (an R&D priority). Globally, biofuel production amounted to roughly 16bn gallons in 2007, up from 10bn gallons in 2005 and 4.8bn gallons in 2000. Even so, the industry still accounts for less than 3% of global transportation fuels. Industry projections suggest production could exceed 87bn gal by 2020.

EXHIBIT 5: GLOBAL BIOFUEL PRODUCTION BY REGION (BN GAL/YEAR), 2007E AND 2020E

(bn gal*) 2007E % 2020E % CAGR North America 6.9 43% 30 34% 12% Europe & Eurasia 2.4 15% 20 23% 18%Asia-Pacific 1.6 10% 30 34% 25% South & Central America 5.1 32% 7 8% 2%Total 16 87 14%

Source: DuPont *NB: 1 gallon = 3.785 liters

To get there, however, the industry expects persistently high energy costs as well as significant government support in the form of mandates, as well as advances in industrial biotechnology. Some of the targeted advances include new, tailored crops with higher starch content (enhancing yield by 3%–5%), advances in enzyme technology to better process cellulose, process improvements to improve thermal efficiency (reducing ethanol costs by as much as $0.20/gal), and, importantly, using co-products from ethanol production to manufacture biodiesel. Other targeted advances are intended to reduce the degree of direct competition between biofuel and traditional crop applications, particularly by shifting away from corn in favor of non-food crops. Clearly, changes in the energy cost outlook, regulatory support, or the pace of technical progress could alter industry forecasts. The industry also needs to put to rest a persistent debate over the energy efficiency of using biofuels.

• At current corn and energy costs, we believe process economics continue to favor producing corn ethanol, at least as a source of carbon.

EXHIBIT 6: CHANGE IN THE FEEDSTOCK LANDSCAPE: COST OF CARBON IN CORN VS. IN CRUDE OIL

Source: Cargill, Industrial Biotechnology. Jefferies & Company, Inc. estimate

Industrial Biotech


$0

$20

$40

$60

$80

$100

$120

$140

$0

$0.5

0

$1.0

0

$1.5

0

$2.0

0

$2.5

0

$3.0

0

$3.5

0

$4.0

0

$4.5

0

$5.0

0

$5.5

0

$6.0

0

$6.5

0

$7.0

0

Corn ($/bushel)

Cru

de O

il ($

/bbl

) Above the line, corn is a cheaper source of carbon

Below the line, oil is a cheaper source of carbon

$0

$20

$40

$60

$80

$100

$120

$140

$0

$0.5

0

$1.0

0

$1.5

0

$2.0

0

$2.5

0

$3.0

0

$3.5

0

$4.0

0

$4.5

0

$5.0

0

$5.5

0

$6.0

0

$6.5

0

$7.0

0

Corn ($/bushel)

Cru

de O

il ($

/bbl

) Above the line, corn is a cheaper source of carbon

Below the line, oil is a cheaper source of carbon


Page 72 of 212

• At 6.5bn gallons, U.S. ethanol production in 2007 contributed less than 3% of the gasoline supply but represented roughly 21% of corn demand and 23% of corn production (the gap due to inventory reductions). In 2008, ethanol production is expected to consume 33% of the U.S. corn crop.

• With 85%–95% of US ethanol sold under 6- to 12-month fixed price contracts or tied to the price of gasoline, the volatility in ethanol spot prices does not directly translate into industry cash flows. As such, the industry has more flexibility to hedge its input costs against visible output prices.

• Ethanol yields are rising. Plants under construction are expected to average 2.8 gal/bushel of corn, up from 2.5 gal/bushel in the 1990s and 2.4 gal/bushel in the 1980s. Improved fermentation processes and higher-starch corn varieties could lift ethanol yields to north of 3.0 gal/bushel by early next decade.

• Long before any theoretical limits are reached, supply constraints are moving ethanol feedstock costs towards equilibrium by driving corn prices and acreage higher. We expect the U.S. and Europe to increasingly import biofuels from Latin America and, eventually, Africa.

• The gap between corn ethanol and cellulosic ethanol appears to be narrowing more quickly than expected, mostly due to the steady upward bias in corn prices. Better enzymes, genetically modified crops, and better plant designs are facilitating this process, and we expect companies to increasingly emphasize feedstock flexibility as a way to reduce feedstock risk.

• Regardless of the feedstock, we remain skeptical that the 85% ethanol fuels will be introduced quickly. While roughly 6 million flex-fuel vehicles in the United States can burn E-85 fuel, the real bottleneck is deploying the necessary infrastructure at gas stations, which usually entails more risk for small operators. Also, in our view the jury is still out as to whether the infrastructure costs for rolling out E-85 will prove prohibitive given rival alternative power technologies (electric, hydrogen). If the United States simply shifted to 10% ethanol content in its fuel supply, however, annual demand could eventually exceed 15bn gal, roughly 3–5bn more than the likely capacity from the corn crop. To achieve such a high target, we expect most of the volume will need to be supplied from cellulosic feedstocks. As such, we do not expect a significant roll-out of E-85 vehicles until the technology for cellulosic-based ethanol is proven to be viable at a commercial scale. Moreover, we believe that liquid fuel/electric cars such as GM’s Volt represent a more compelling trend, given the opportunity to leverage the existing electrical grid without the infrastructure issues that bedevil E-85.

• While most of the focus in recent years has been on corn ethanol, favorable tax incentives have also encouraged investment in biodiesel, a biodegradable and non-toxic alternative to conventional diesel made using vegetable oils and animal fats. Industry capacity, at 2.2bn gal/year, represents only a quarter of ethanol installed capacity: the plants currently proposed would increase U.S. capacity to 3.4bn gal/year by 2010, still roughly a third of the expected ethanol capacity base. Actual production levels are significantly lower, with total U.S. production expected to reach 450m gallons in 2007, rising to 600m by 2013 (4.9% CAGR) or 1,000m gallons (14.2%) if the 2007 RFS mandate is implemented. Feedstock supplies, however, are increasingly strained, with energy applications representing more than 17% of domestic soybean oil demand in 2007 (vs. 68% of rapeseed oil in Europe). As a result, biodiesel producers using human-grade soybean oil and palm oil are for the most part unprofitable even with the government subsidies.

• Waste feedstocks for biodiesel, such as inedible tallow and yellow grease, have better economics, but supply is limited. Indeed, even if the industry can convert all the available grease, soybean oil, and tallow into biodiesel, it would only be large enough to displace 5% of current diesel demand (vs. 0.1% last year). Any push to further adopt biodiesel will hinge on the availability of feedstock imports (e.g., palm oil), which should entail structurally higher feedstock prices than domestic waste products. We estimate that using palm oil rather than waste products adds $1.20/gal.

• Biofuels demand has contributed to the rise in corn, palm, soy, barley, wheat and rapeseed prices in recent years. Within three to five years, demand for cellulosic ethanol could start to have an impact on wood prices as well. The most powerful lever to offset this is to increase crop production yields. In the U.S., for example, rising corn yields meant that the U.S. could maintain the pace of corn exports even as it ramped up domestic ethanol production.

Industrial Biotech



Page 73 of 212

EXHIBIT 7: U.S. CORN EXPORTS AND CORN USED FOR ETHANOL PRODUCTION

0

2

4

6

8

10

12

14

1990 1995 2000 2005 2010 2015

U.S. corn: Domestic use, ethanol, and exports

Billion bushels

Total domestic use

Exports

Ethanol

0

2

4

6

8

10

12

14

1990 1995 2000 2005 2010 2015

U.S. corn: Domestic use, ethanol, and exports

Billion bushels

Total domestic use

Exports

Ethanol

Source: USDA

• Ethanol has entered a consolidation phase. In 2005–1H06, ethanol shares were still in the “scarcity” phase, when first movers in ethanol fetched a premium due to the prospect of large near-term windfalls as well as the relative scarcity of investment vehicles to play the favorable newsflow. As more ethanol producers raised capital, the sector moved into a phase based on differentiating relative cost positions. Sector valuations compressed, and projects have started to be cancelled. The industry now has 7.8bn gpy of capacity (vs. a mandate of 5.4bn gpy for 2008), and new projects on track to lift capacity to 13.4bn gpy (down from 14bn gpy six months ago). To improve returns, the sector has started to consolidate, with a focus on producers in the middle of the cost curve. With relatively few large-scale plants available, however, consolidators could be hard-pressed to acquire facilities below replacement cost ($1.50–$2/gal). We expect smaller producers at the higher end of the cost curve to be displaced.

• Government mandates sometimes have perverse consequences. The $0.51/gal blenders credit, for example, is not always passed on to the ethanol producers due to distribution bottlenecks. The proposed Farm Bill, moreover, would cut the blenders credit to $0.45/gal (a $1bn cut in the subsidy), which would be partly offset by a new $1.01/gal ($500m) subsidy for cellulosic ethanol.

Winners and losers. Production economics, in our view, are probably the most significant guide to producers’ viability and profits, which in turn drive shareholder value. While the entire industry is exposed to the issues related to near-term volatility in key feedstock costs (corn, natural gas) and the price of ethanol and gasoline, and longer-term risks from the capacity build, investors will likely differentiate both on near-term advantages due to successful hedging or feedstock sourcing (i.e., using EPS momentum as a proxy) as well as sustainable differences in cost structure. For example, larger facilities (economies of scale), flexible plant design (to be ready to switch to other feedstocks), lower collection and distribution costs (e.g., located on one of the coasts and buying/selling locally), or more flexible business models (e.g., investments in storage capacity and marketing services) should be able to mitigate the impact of higher corn prices on margins, and consequently receive a more attractive relative valuation in 2008–2009. Accelerating investment, subsidies and grants for cellulosic ethanol should bode well for suppliers of enabling technology and products (e.g., seeds and enzymes). On the other hand, smaller, older plants, particularly those located in regions where significant capacity additions are under way, companies without sophisticated hedging practices, and companies with stressed balance sheets are the most likely losers over the next few quarters. Companies with no plants in the ground and business plans predicated on plants starting up in 2009–2010 will likely be at risk as well.

Industrial Biotech



Page 74 of 212

Government Policies Shape Global Landscape for Biofuels In recent years, governments around the world have introduced initiatives to encourage the adoption of biofuels as a renewable alternative to fossil fuels and nuclear energy. Regulatory changes remain the dominant driver for the sector going forward.

In the United States, the regulatory landscape is in flux. Federal subsidies at various levels have been in place since 1978, supplemented by state subsidies. As recently as late 2007, momentum appeared to favor increasing ethanol subsidies. The U.S. had a $0.51/gal subsidy for blending ethanol as a fuel additive, while biodiesel receives a $1/gal subsidy if made from virgin oil and $0.50/gal if made from recycled oil such as cooking grease. The U.S. also maintained a $0.54/gal tariff on ethanol imports in order to restrict competitive pressure from Brazil, though at current sugar and shipping costs Brazil’s ability to import into the United States appears fairly limited. In 2006, the phase-out of MTBE (a fuel additive made from petrol and alcohol) effectively increased ethanol demand by roughly 2bn gal/year. Only a few months ago, the federal government rolled out new mandates as part of the 2007 Energy Act. States had aggressive mandates in place as well: Missouri, for example, had a 10% ethanol mandate (300m gpy of ethanol demand). All in, industry observers estimated ethanol receives direct and indirect subsidies of $1.05–$1.40/gal, or 45%–60% of the current market price. EXHIBIT 8: RENEWABLE FUEL STANDARD FROM THE 2007 ENERGY ACT (TAKES EFFECT JAN. 1, 2009)

Year Corn Ethanol Cellulosic Biodiesel Other Biofuel* Total % YoY 2008 9 0 0 0 9 2009 10.5 0 0.5 0.1 11.1 29%2010 12 0.1 0.65 0.2 12.95 18% 2011 12.6 0.25 0.8 0.3 13.95 9%2012 13.2 0.5 1 0.5 15.2 11% 2013 13.8 1 - 1.75 16.55 11%2014 14.4 1.75 - 2 18.15 13% 2015 15 3 - 2.5 20.5 17%2016 15 4.25 - 3 22.25 12% 2017 15 5.5 - 3.5 24 11%2018 15 7 - 4 26 11% 2019 15 8.5 - 4.5 28 10%2020 15 10.5 - 4.5 30 11% 2021 15 13.5 - 4.5 33 14%2022 15 16 - 5 36 12%

Source: EPA *Could include cellulosic ethanol and biodiesel, as well as other biofuels such as biobutanol.

Election year politics, which had once appeared to be a “guaranteed” catalyst for the sector as candidates appeal to Iowa, has turned into a headwind. Texas and Connecticut have appealed for an EPA waiver of the ethanol mandate, Missouri is considering reversing its new 10% ethanol mandate, and Presidential candidate McCain and 22 other Republican senators have appealed to the EPA to waive or amend the ethanol mandate passed in 2007. The perception of heightened political risk could affect the discount rates applied to biofuels until after the next election, at least.

Brazil, the world’s largest producer of ethanol, has made the most effort to use ethanol as a substitute for fossil fuels. In response to the 1973–1974 oil shock, Brazil launched the Proalcool program, which initially provided incentives for producers and tax rebates for consumers. Currently ethanol use is not subsidized, and hydrated ethanol is competitive in the domestic market at 60%–70% of the price of E10 gasoline. Since 2003, Brazil has displaced approximately 40% of its oil consumption with ethanol, and flex-fuel cars now represent roughly 75% of annual new vehicle sales (vs. 4% in 2003). Brazil has an advantage due to lower per capita energy consumption and lower labor costs, but this shift has also been due to government discipline. The government mandates a 20%–25% blend of ethanol with gasoline, sets lower excise taxes on ethanol storage, and protects the domestic producers with a 20% duty on imports. Brazil has also introduced a biodiesel mandate: 2% by 2007 (800M l), 5% by 2013 (2bn l), and 20% by 2020 (12bn l).

In Europe, the focus has been on biodiesel, which represents nearly 80% of the biofuel market. Germany, France, and Italy have taken the lead. With more than 15,000 biodiesel filling stations, Germany alone accounts for more than half of global biodiesel production and consumption. With transportation contributing an estimated 21% of EU greenhouse emissions, and increased concerns over volatile oil prices, the European Commission has set targets for renewable fuels of 2% of all transport fuels in 2005 (in practice less than 1.4% in 2005), rising to 5.75% by 2010. This compares with an estimated 0.6% in 2003. Germany exempts biofuels and biomass-based heating oils from duties until 2009. France exempts 1.2M tpy of biofuels from excise taxes and has a progressive tax rate tied to the

Industrial Biotech



Page 75 of 212

amount of biofuel blended into a transport fuel. Spain has adopted the EU’s 5.75% target by 2010, and has also set a target of 2.2M tpy of biofuels (in tons equivalent to petrol) by 2010, versus 0.2M tpy in 2004. Sweden fully exempts biofuels from excise duties and targets 3% of total transport fuels. In 2005, European ethanol production amounted to roughly 1M tonnes, with capacity on track to triple by 2007. Biodiesel capacity is expected to exceed 4M tonnes by the end of 2006.

In Asia, Malaysia is planting more than 600,000 ha of palm oil plantations to support displacing 5% of diesel consumption with palm oil-based biodiesel in 2007. Longer term, Malaysia has set a target of providing 10% of global biofuel supply via palm oil. India has introduced a 5% blend mandate for several states, Thailand has provided tax incentives to support 10% blends, and Australia has set tax incentives through 2015 (with a step down in support after 2011). Japan targets using 500M liters of biofuels by 2010. Finally, China is investing more than $500M/year in agricultural biotech, with an eye to sponsor domestic companies such as Weiming and Biocentury that will develop and market inexpensive genetically modified crops. China has targeted a 10% blend of biofuels by 2020, which implies demand for 23M tpy of biofuel. The five largest Chinese ethanol producers have total capacity north of 1.5M tpy, 80% based on grains, 10% from sugar, 6% from paper pulp waste, and the rest from ethylene. China’s planned biodiesel and ethanol capacity additions imply 11M tpy of installed capacity by 2010. Importantly, China has emphasized shifting new biofuels capacity away from feedstocks that humans consume as well, in sharp contrast to the US policy of ignoring the “food vs. fuel” conflict.

EXHIBIT 9: SUMMARY OF GLOBAL ETHANOL BLENDING MANDATES

Country Ethanol Mandate U.S. 15bn gpy by 2015. $0.51/gal blender's credit probably cut to $0.45/gal in Farm Bill.

New $1.01/gal cellulosic ethanol credit likely. $300/t subsidy for 99% biodiesel exports

Brazil 20%-25% Canada 5% ethanol or biodiesel by 2010

E.U.

10% by 2010 France: 5.75% by 2008, 10% by 2010

Germany: 3.6% ethanol by 2015, 4.4% other biofuels Total EU subsidies for biofuels reached 3.7bn euros in 2007.

Lithuania 3%-7% Poland 3.45% biofuels Argentina 5% by 2010 Thailand 10% in Bangkok India 5% China 10% in 5 provinces Philippines 5% in 2008, 10% by 2010 Bolivia 25% by 2013 from 10% currently Colombia 10% in all cities with more than 0.5m inhabitants Venezuela Phasing in 10%


Key Investment Themes: Focus on Production Costs, Disruptive Technologies

Whereas ethanol production using corn or sugarcane uses basic yeast-based fermentation processes, we believe the rapid progress in biology is likely to deliver step-changes in technology over the next decade. The benefits of industrial biotechnology for agriculture are the most developed, and we believe opportunities in energy storage, bioconversion of heavy-carbon feedstocks, materials production, and carbon sequestration are only starting to be addressed. Currently, these opportunities are being blended together, which will likely lead to inefficiencies in the value chain. For investors interested in near-term developments, we would recommend a focus on companies that can help optimize key parts of the value chain, particularly feedstocks, conversion yields, and the ease of dropping alternate biofuels into the existing distribution infrastructure. In the long run, however, we expect differentiation in feedstocks to be accompanied by differentiation in co-products and plant design. This should create a premium for either platform companies that have the skill set to help other companies integrate plant design with metabolic engineering of the fermentation microbes, or companies that do a better job running integrated production facilities to improve co-product economics. Enablers that help these companies create their own value chains, rather than piggy-backing on the existing agricultural or petrochemical value chains, should be attractive longer-term plays as well.

Industrial Biotech



Page 76 of 212

With this in mind, key investment themes include the following:

• Biodiesel vs. ethanol. On a broad-brush basis, we favor biodiesel over ethanol. Biodiesel lacks the supply-chain issues that bedevil ethanol, biodiesel production processes typically have more flexible feedstocks, and biodiesel demand represents a much smaller portion of demand for those feedstocks. In either case, however, we expect to see new entrants compete with more innovative and thermally efficient plant designs over the next few years. Plant designs that generate less waste, with a higher yield, should be able to provide a sustainable cost advantage and higher returns on capital. This process of plant optimization, however, is a recipe for pushing existing capacity towards the higher end of the cost curve, and as the industry matures we expect ethanol shares to trade below replacement costs to reflect this ongoing cycle of innovation. This implies significant risk for consolidators given the current premiums being paid for ethanol capacity.

• Alternatives to yeast and corn for ethanol production. To date, ethanol production has focused on using yeast to ferment corn or sugarcane. Genetically modifying yeast is improving ethanol yields (on a gal/bushel basis), and modified enzyme cocktails have further improved ethanol economics. To make cellulosic ethanol economical, however, research is broadening away from yeast towards genetically modified bacteria in order to generate some of the enzymes organically, with some additional cellulases added as part of the conversion process. Others are starting instead from bacteria which produce their own cellulase. We are also seeing more investment in gasification processes that use inorganic catalysts (and potentially, microbial catalysts) rather than (exogenously supplied) enzymes to improve reaction efficiency. Besides the range of production processes, ethanol producers need to ensure that the feedstock is environmentally sustainable: that is, it does not damage soil quality or the water table to the point of becoming unprofitable. Non-sustainable processes, in our view, will likely run into increasing regulatory obstacles in the coming years as consumers and governments become more sophisticated in their evaluation of environmental impacts.

• Biochemicals vs. biofuels. Similarly, we believe that biochemicals and biomaterials are better positioned to deliver returns across the cycle than biofuels. Again, in biofuels we see several plausible technologies likely to compete for investor attention by 2009–2010, and we expect new entrants with better plant design, co-product integration, and fermentation (biotransformation) processes to increase the level of competition in each product line. In biochemicals and biomaterials, we see a better opportunity to establish a competitive advantage in IP that will enhance the scarcity value of the product. Moreover, by selling products that compete on value rather than a cost-plus commodity model, producers of specialty biomaterials should maintain some degree of insulation from volatile feedstocks. Finally, we believe we are very early in the cycle of government incentives for biomaterials, whereas government incentives for biofuels are already well established—and well anticipated by share prices.

• Strong IP. In particular, we would focus on companies with unique IP in the industrial biotechnology triangle of genetically modifying feedstock crops, discovering and modifying microbes to optimize the fermentation process, and developing cocktails of enzymes to facilitate multi-step, multi-phase and variable temperature processes. Given the rapid pace of development of industrial biotechnology, these areas should have more sustainable barriers to entry than companies that rely on competitive advantages based on plant engineering, logistics, or prowess at hedging financial or commodity risk. Even so, companies that specialize in gene discovery, enzyme optimization, plant design and construction, or product distribution still face significant risk of commoditization, particularly as larger sums of capital are allocated to industrial biotechnology R&D.

• Integration vs. platform companies. On a similar broad-brush basis, we expect the market to favor companies that integrate scientific platforms, design capabilities, production, and channels to market over platform companies. Integrating facility design, channels to end market, and genetic modification of feedstocks and fermentation processes will emerge as a key driver of higher returns on capital, in our view, because many of the essential business choices are “chicken and egg” questions. Which crop to grow, how best to grow it or hedge the supply from a third party, how to ensure the crop is environmentally sustainable, how to process it, what to turn it into, how to process the co-products, where to locate the plant, whether to optimize the process or retain flexibility, how best to get the products to market—the answers to each of these questions has an impact on the others, and on the long-term viability of the facility, and in many cases they are being evaluated without reliable precedents. At the current stage, an integrated approach should provide synergies in terms of tailoring fermentation microbes to the production process, and vice versa, as well as being able to accelerate the optimization of biomaterials to suit consumer requirements. As the industry matures, companies can spin-out non-core but capital-intensive processes, much as the refineries have outsourced their hydrogen supply in exchange for higher reliability of supply. In the meantime, as early stage companies are typically capital constrained, we expect a wide range of business models to evolve to bring down the aggregate cost of capital for the new technology.

Industrial Biotech



Page 77 of 212

• Keep an eye on the majors. The oil majors are increasingly involved in biofuels, both as transportation fuels and, at least anecdotally, as feedstocks for chemicals. ConocoPhillips and Tyson, for example, have announced an alliance to produce and market diesel fuel made from pork, poultry, and beef fat with an estimated $150–$175M facility in the US producing 175M gpy of biodiesel by 2009. ConocoPhillips also produces diesel from soybean oil at a plant in Ireland, and developed its fat-based diesel in Ireland as well. Most recently, it announced an alliance with ADM to develop biocrude, a more flexible intermediary, out of crops, wood and waste products. Chevron, meanwhile, is investing in a 100M gpy soybean-based biodiesel plant in Galveston, Texas, as well as in cellulosic ethanol. Outside the US, BP is investing in biobutanol (with DuPont), ethanol, and biodiesel (albeit with some controversy about the specifications). PetroChina, meanwhile, has drawn attention with its plans to produce 667M gpy of non-grain ethanol by 2010 and 60M gpy of biodiesel. On the infrastructure side, while the U.S. debate continues about the feasibility of ethanol infrastructure, Petrobras is building in Brazil a $235M dedicated ethanol pipeline that will run approximately 800 miles from the Goias state to a San Paulo refinery and then to the port of Sao Sebastiao. Competition from the oil majors, for both end markets and subsidies, could prove disruptive on a regional basis, particularly for producers that have older, smaller facilities that might face adverse transportation costs.

• BP and Verenium announced a strategic partnership in August, focused on the commercialization of cellulosic ethanol; in the first stage, BP pays to support ongoing cellulosic ethanol R&D and, in the second stage, the two companies intend to create a JV to build cellulosic ethanol facilities in the US. An earlier collaboration (May) announced between DuPont and Danisco (Genencor) aims to build an integrated technology platform and IP/technology commercialization model that could hand a developed ‘steel, set-up and operating’ solution to a range of customers. Their ambition, too, is to deliver the “lowest cost process” for cellulosic ethanol. However, timelines look further out than those of Verenium, who confirmed its own target for commercial plant/partnership underway by 2009/2010 (and first commercial scale production in 2011). Current DuPont/Danisco timelines involve regional pilot facilities by 2009 (in the US) and a demonstration plant (operational) in 2012 — although the opportunity to beat those dates is likely to be on the wish list of both companies.

Another related investment theme would be opportunities to alleviate key bottlenecks in the supply chain, indirectly exploiting opportunities created by government mandates:

• Distribution. Ethanol is primarily distributed by truck, rail, or barge. Ethanol is hydrophilic, which can complicate using a pipeline. This property also provides an incentive for dedicated transport fleets (less time spent cleaning containers), which should provide an opportunity for transportation companies to improve their utilization rates. According to media reports, approximately 35%–40% of the new orders for railcars are for cars that can ship ethanol. Given the tight operating conditions in the railcar industry, however, we expect many of these new cars to be delivered in 2H07–2008. This implies that ethanol producers that have some degree of control over their distribution infrastructure will be best positioned over the next couple of years, while the rest could face significant cost inflation. At the same time, while rail capacity is a critical issue for ethanol producers, it is, in our view, immaterial for the railroads, as ethanol shipments represent only about 1% of total railcar loadings.

• Storage. As noted above, ethanol tends to separate from gasoline over time. As a result, it cannot be shipped in pipelines, and needs to be blended at distribution terminals rather than at the refinery. This implies the need for a significant increase in storage capacity over the next couple of years to accommodate the doubling in ethanol capacity expected by 2010. Companies that have significant storage capacity already on hand, such as The Andersons, will likely receive a premium for these assets regardless of the long-term economics of ethanol versus substitute biofuels.

• Engines. Most cars in the United States cannot run on fuel that is more than 10% ethanol. Only about 6 million vehicles in the U.S. can run on E85, fuel that is 85% ethanol. The retrofit, however, is relatively inexpensive, with industry observers estimating it at about $100/car (mostly for fuel lines, fuel tanks, and sensors). Companies that supply these components should see solid demand trends over the next three to four years, at least.

• Feedstocks. Even with its current fairly limited volumes, ethanol demand already represents roughly 20% of US corn demand. Multiple industry observers, including DuPont, have estimated that by the time U.S. ethanol demand exceeds 5bn gallons, the industry will be nearing the limit of how much corn it can use without having an exaggerated impact on corn prices. Furthermore, the DoE has estimated that the United States faces a longer-term constraint on the amount of corn that can be switched from food to ethanol, limiting ethanol volumes at roughly 12bn gal—a level the industry could reach as early as 2009. The next step will likely be a shift towards using waste cellulose, such as corn and wheat stalks or other vegetable matter. In the long run, we expect the industry to move towards dedicated energy crops, where the ethanol production can be integrated with chemical production much as in a traditional refinery. (For more details on the long-term potential, please see our April 24, 2006, primer, “Industrial Biotechnology”.) This should reduce the direct pressure on corn prices and further differentiate biorefinery economics from the petrochemical value chain.

Industrial Biotech



Page 78 of 212

EXHIBIT 10: POTENTIAL TRAJECTORY FOR ETHANOL INDUSTRY

2005 Interim Long-Term Feedstock Starch Waste cellulose Cellulosic energy crops (3x

yield per acre) US Volume (bn gallons) 4 20 30-200 % fossil fuel displaced 2% 10% 15%-100% CO2 reduction 1.80% 9% 14%-90%

Process

Starch fermentation,

limited cellulose processing.

Enzymatic hydrolysis, Cellulases, Single Sugar metabolisms, multiple

microbes, energy crops, more residue processing

Designer cellulosic energy crops, integrated processing, optimized

pretreatment, various/single microbe(s) capable of hydrolyzing cellulose and

producing ethanol, carbon sequestration through plant partitioning

Deployment Large, central processing Large, central processing Distributed or centralized

Energy yield 14% 37%-plus Est. $/gal gasoline equivalent $3.78 $1.62

Current mfg. costs 50% > starch $0.91

Source: DOE

EXHIBIT 11: THEORETICAL YIELD OF ALTERNATE FEEDSTOCKS (ETHANOL GAL/TON OF FEEDSTOCK)

Corn Grain 124.4 Corn Stover 113.0 Rice Straw 109.9 Cotton Gin Trash 56.8 Forest Thinnings 81.5 Hardwood Sawdust 100.8 Bagasse 111.5 Mixed Paper 116.2

Source: DOE

• Land could also emerge as an important bottleneck for the industry, implying upside potential for companies with real estate holdings close to new production facilities. To put this in perspective, the United States has a total surface area of roughly 1.9bn acres (in the 48 contiguous states), but only about 400 million acres of cropland. Switchgrass is grown on only about 30 million acres of land covered by the Conservation Resource Program. Already the increased demand for biofuels has translated into higher prices for crop land. According to media reports, in 2006 the average price of farm land rose 15% in the United States, 27% in Argentina, and 10% in Australia. In Idaho, crop land prices rose 35%.

According to an NRDC study conducted in 2004, more than 1.7bn acres would need to be shifted to biofuels if current technology were to displace U.S. gasoline consumption (base case 289bn gal in 2050, up from 137bn gal in 2004). The NRDC believes this could be reduced to a more manageable requirement, below 50 million acres, if several (admittedly ambitious) innovations are delivered, including the following:

• Raise fuel efficiency to 50 mpg and implement new mass transit initiatives which could reduce gasoline demand by almost 40%.

• Improve crop yields as well as the conversion efficiency for converting cellulose to ethanol. • Produce biodiesel from the waste products. • Convert soybean production for animal feed to switchgrass and collect corn stover for conversion to fuel.

Industrial Biotech



Page 79 of 212

EXHIBIT 12: LAND NEEDED TO MEET ENERGY NEEDS IN 2050 WITH BIOFUELS

Gasoline demand Switchgrass Yield Conversion Efficiency Land Production and efficiency (bn of gals of gas equiv) (dry tons/acre/year) (gals gas equiv/ ton) (m acres) Status Quo In 2050 289 5 33 1753 Smart Growth & Efficiency 108 5 33 657Increase Conversion Efficiency 108 5 69 313 Biofuels Coproduction 108 5 77 282Increased Switchgrass Yield 108 12 77 114

Animal Feed From Switchgrass (73m acres, 50%-100% converted) 41-77 Corn Stover (323m tons, 75% collected) 21-88 Conservation Reserve Switchgrass (30m acres, 33%-50% converted) 6-48

Source: NRDC

How Much Is Capacity Worth?

As a rapidly emerging industry with volatile supply/demand dynamics, biofuels have experienced wide swings in valuations. Short-term traders frequently tie ethanol share performance to near-term trends in the price of ethanol, with gasoline prices, corn prices, and natural gas prices having a more limited impact on share volatility. The simplest fundamental approach, perhaps, is to draw an analogy to refineries, and place a comparable 5x–6x EBITDA on normalized estimates (e.g., once projected plants come onstream in 2009E or 2010E).

The nature of the biofuels market and the capital intensity of the industry have also led to widespread use of asset-based valuation models. In these models, the different biofuel producers are put on a common valuation basis by looking at the ratio of their enterprise value to their installed or expected capacity.

Capital costs for new ethanol capacity have risen to roughly $2–$2.25/gal, up from $1.50/gal in 2003–2004. Some of the increase has been driven by design changes, but the lion’s share reflects higher metal prices and higher labor costs as the plant builders such as Lurgi and Fagen are running full out and have significantly extended their construction times for new plants, such that any new contracts are likely achievable only in 2009–2010.

• In the near term, while the industry cost curve remains relatively flat, we expect ethanol companies to trade above their replacement cost, or else become candidates for consolidation.

• In the long run, however, we expect improvements in plant design to lead to newer plants being consistently at the low end of the cost curve. As a result, equities for companies with older facilities will likely end up trading below their replacement value on an asset basis. In other mature capital-intensive industries such as offshore drillers and chemicals, discounts of 10%–15% to replacement asset values are typical and 30%–40% plausible at cyclical troughs.

A more nuanced approach, in our view, is to value ethanol capacity based on models of process economics. We give examples of such models in the next few pages. These models can be used in a couple of different ways. First and foremost, they can be used to assess what the long-term value of biofuel production might be under various market conditions. For ethanol, such an analysis would go beyond simply looking at the “crush spread,” or the spread between ethanol prices and the requisite corn input costs, and takes into account catalyst costs, recovery of capital, natural gas prices, and other cost inputs.

Perhaps more important, the models provide a way to assess the “full cycle” value of biofuel capacity. Our key assumption is that, over time, the prices of oil-based and bio-based fuels will settle into an equilibrium such that biofuel producers generate long-term rates of return comparable with other commodity industries, or slightly below their cost of capital (i.e., 10% EBIT margins, 6%–7% after-tax ROIC). Simply put, in an industry where the marginal producers price their product at the cash cost of production, we expect producers to recover their capital costs, but not their initial fixed costs. Higher returns would provide an incentive to add additional capacity, putting upward pressure on input costs (e.g., corn in the case of ethanol, or rapeseed in the case of European biodiesel).

Such a valuation approach lends itself to a “base plus” analysis. After establishing a benchmark full-cycle valuation, one can then analyze the returns implied in current market prices to see whether likely favorable or unfavorable factors are adequately discounted. For example, one can infer the amount of “windfall” profits the producers are expected to generate before new competitor capacity brings returns down to full-cycle levels. Alternately, one can set equilibrium pricing based on the highest cost of production (e.g., cellulosic ethanol), and then explore the implications for fair value for the other producers. Our equilibrium analysis supports $1.50–$1.75/gal as a long-term fair value, or roughly 80%–90% of replacement costs. Notably, in the near term, cash margins are below likely equilibrium levels, which should slow the capacity build out in the industry.

Industrial Biotech



Page 80 of 212

Importantly, such an equilibrium analysis would be an estimate of theoretical fair value for “typical current capacity”, excluding hedging and financing costs. Capacity to be built for production in 2009, for example, would need to be discounted appropriately to reflect the fluid state of the industry (e.g., 15%–20% discount rate). Moreover, for actual facilities, the NPV would necessarily change due to that facility’s process economics, feedstock sourcing costs, co-products, logistical advantages or disadvantages, and operating rates, among other factors. Finally, our models exclude any estimate for SG&A, corporate overhead, or financing costs, which would pertain to equity valuations, and by extension take-out valuations.

State of the Ethanol Market

Ethanol is an alcohol that can serve as a fuel as well as a substitute for the oxygenate MTBE. It has emerged as one of the highest-profile alternative energy plays due to favorable economics, particularly in 1H06, and high-profile political support. Ethanol is derived from sugar, which in turn can be generated from corn (in the U.S.), sugarcane (in Brazil), wine (Sweden), wheat, beets, or even cellulosic waste products such as stems and stalks. Ethanol was first used in the US in the 1820s, and was an automotive fuel by the start of the 20th century. While the U.S. government has subsidized ethanol production since the 1970s, the recent combination of high gasoline prices, the US phase-out of MTBE (2bn gal of incremental demand, or 50% of 2005 production), and process improvements made ethanol production economical, and until recently investors could be confident that government mandates would continue to support the expansion of the industry.

Rapid growth unlikely given pressure on mandates

Ethanol production in the U.S. has grown rapidly, more than doubling since the late 1990s due to the federal and state mandates discussed above. As a result, by 2005 U.S. production capacity matched that of Brazil. At least 61 new projects are contemplated, amounting to 5.5bn gal/year. Given current capacity of 7.8bn gal/year, the proposed capacity expansions, if completed, would bring total installed capacity to 13.4bn gal/year—equal to the federal mandate for 2012. With the mandates unlikely to move higher, the base case for the industry should be a steady erosion in operating rates into the end of the decade, with the market tightening again in 2011–2013. Moreover, once ethanol production exceeds the mandated usage requirements, we expect ethanol to shift to competing with gasoline on an energy content basis—i.e., at a 35% discount to gasoline prices.

EXHIBIT 13: US ETHANOL PRODUCTION (M GAL AND % YOY, 1980–2012E): DECELERATING

2,000

4,000

6,000

8,000

10,000

12,000

14,000

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

E

2010

E

2012

E

-30%-20%-10%0%10%20%30%40%50%60%70%

% YoY m gallons, LHS

2,000

4,000

6,000

8,000

10,000

12,000

14,000

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

E

2010

E

2012

E

-30%-20%-10%0%10%20%30%40%50%60%70%

% YoY m gallons, LHS


Industrial Biotech


0


Page 81 of 212

EXHIBIT 14: US ETHANOL YEAR-END CAPACITY (BN GPY, LHS) AND % UTILIZATION RATE (RHS) BASED ON 2007 MANDATE

1.7 1.9 2.3 2.7 3.1 3.6 4.3 5.57.9

9.711.5

13.4 13.4 13.4

84% 85%75% 79%

91% 94% 92% 89%82%

93% 91% 89% 94% 98%

5

10

15

20

25

30

35

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

E

2009

E

2010

E

2011

E

2012

E

0%

20%

40%

60%

80%

100%

120%

01.7 1.9 2.3 2.7 3.1 3.6 4.3 5.5

7.99.7

11.513.4 13.4 13.4

84% 85%75% 79%

91% 94% 92% 89%82%

93% 91% 89% 94% 98%

5

10

15

20

25

30

35

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

E

2009

E

2010

E

2011

E

2012

E

0%

20%

40%

60%

80%

100%

120%

0


Due to this rapid expansion of capacity, the United States has become the largest producer of ethanol in the world. The U.S. market share rose 11% to 50% in 2007, and we expect U.S. share of global capacity to continue to increase over the next few years, given the doubling in capacity expected by the end of the decade. Persistent high corn prices, particularly if not offset by higher gasoline (and consequently ethanol) prices, could lead to project cancellations or delays, particularly of the smaller facilities that will likely end up at the high end of the cost curve.

EXHIBIT 15: WORLD ETHANOL CAPACITY (M GAL AND % OF TOTAL): 2007 VS. 2005

2007 (m gpy) % 2005 % % CAGR U.S.A. 6,499 50% 4,264 35% 23.5% Brazil 5,019 38% 4,227 35% 9.0% EU 570 4% 516 4% 5.1% China 486 4% 1,004 8% -30.4% Total 13,102 100% 12,150 100% 3.8%

Source: RFA

The surge in US capacity is ironic, as US ethanol had higher production costs even before the recent run in corn prices. The following two tables present estimates from the USDA and OECD, which echo anecdotal estimates.

EXHIBIT 16: ESTIMATED COSTS OF DIFFERENT BIOFUEL PROCESSES, 2004

Product Feedstock Energy Processing Co-Product Credits Net cost $/t $/gal $/t $/gal $/t $/gal $/t $/gal $/t $/gal

EU Biodiesel (Vegetable Oil) 608 1.82 53 0.16 86 0.26 62 0.19 685 2.05 EU Ethanol (Wheat) 448 1.34 73 0.22 431 1.29 228 0.68 725 2.17 EU Ethanol (Sugar Beet) 381 1.14 73 0.22 358 1.07 105 0.31 707 2.12 US Ethanol (Corn) 245 0.73 80 0.24 130 0.39 90 0.27 365 1.09 Brazil Ethanol (Sugar Cane) 163 0.49 0 0.00 114 0.34 0 0.00 276 0.83 Gasoline (cash cost) 0.40-0.60

Source: OECD

Industrial Biotech



Page 82 of 212

EXHIBIT 17: ESTIMATED COSTS OF DIFFERENT ETHANOL FEEDSTOCKS, 2003-2005 ($/GAL)

Feedstock US US US US US US Brazil EU Corn Wet

Milling Corn Dry Milling

Sugar Cane

Sugar Beets Molasses

Raw Sugar

Sugar Cane Sugar Beets

Conversion factor (gal/t) 98.21 98.21 19.5 24.8 69.4 135.4 19.5 24.8 Cost structure ($/gal)

Feedstock (net of co-products) 0.40 0.53 1.48 1.58 0.91 3.12 0.30 0.97

Processing 0.63 0.52 0.92 0.77 0.36 0.36 0.51 1.92 Total cash cost 1.03 1.05 2.40 2.35 1.27 3.48 0.81 2.89 Depreciation (40 m gpy facility) 0.12 0.12 0.16 0.16 0.10 0.10 0.16 0.16 Total cost 1.15 1.17 2.56 2.51 1.37 3.58 0.97 3.05

Source: USDA

Margin outlook EXHIBIT 18: US ETHANOL PRICES ($/GAL), 1998-2008

Source: Bloomberg

US ethanol prices have been volatile over the past three years, with upwards price pressure through early 2006 driven by both rising gasoline prices (a function of higher crude oil prices) and, in the first half of 2006, the phase-out of MTBE. Markets with infrastructure bottlenecks saw the sharpest rise, with spot prices in New York exceeding $5/gal in June even as prices in Iowa remained below $4/gal. New capacity coming onstream and falling oil prices led to a sharp retracement (more than 50%) in ethanol prices to less than $2/gal. As a result, ethanol prices in the spot market fell to roughly $1.70/gal in late 2007 and then recovered to the $2.50/gallon level. Importantly, 85%–95% of US ethanol production is sold under six- to 12-month fixed price contracts or tied to the price of gasoline.

Meanwhile, corn ethanol production is expected to consume 33% of U.S. corn production this year, a function of ethanol capacity expansions and a YoY decline in corn acreage after a strong 2007 planting year. This compares with roughly 22% in 2007, 18% in 2005 and only 8% in 2001. Approximately 8% of the demand for the U.S. corn crop is related to the production of 2bn gallons of ethanol to substitute for MTBE after MTBE was phased out in 2006. Coupled with rising fertilizer costs, this has been a recipe for sharply higher corn prices and consequently margin pressure for ethanol producers.

Industrial Biotech


$0

$0.50

$1.00

$1.50

$2.00

$2.50

$3.00

$3.50

$4.00

Sep

-97

Mar

-98

Sep

-98

Mar

-99

Sep

-99

Mar

-00

Sep

-00

Mar

-01

Sep

-01

Mar

-02

Sep

-02

Mar

-03

Sep

-03

Mar

-04

Sep

-04

Mar

-05

Sep

-05

Mar

-06

Sep

-06

Mar

-07

Sep

-07

Mar

-08

Sep

-08

$0

$0.50

$1.00

$1.50

$2.00

$2.50

$3.00

$3.50

$4.00

Sep

-97

Mar

-98

Sep

-98

Mar

-99

Sep

-99

Mar

-00

Sep

-00

Mar

-01

Sep

-01

Mar

-02

Sep

-02

Mar

-03

Sep

-03

Mar

-04

Sep

-04

Mar

-05

Sep

-05

Mar

-06

Sep

-06

Mar

-07

Sep

-07

Mar

-08

Sep

-08


Page 83 of 212

EXHIBIT 19: RISING OIL PRICES DRIVING INCREASING FERTILIZER COSTS AND OTHER AGRICULTURAL INPUTS

0

50

100

150

200

250

300

1990 1992 1994 1996 1998 2000 2002 2004 2006

Crude oil, natural gas, and nitrogen-based fertilizer prices move together

Nitrogen fertilizer

Crude oil

Source: Producer Price Indexes, U.S. Department of Labor, Bureau of Labor Statistics.

Natural gas

Producer Price Indexes, 1992=100

0

50

100

150

200

250

300

1990 1992 1994 1996 1998 2000 2002 2004 2006

Crude oil, natural gas, and nitrogen-based fertilizer prices move together

Nitrogen fertilizer

Crude oil

Source: Producer Price Indexes, U.S. Department of Labor, Bureau of Labor Statistics.

Natural gas

Producer Price Indexes, 1992=100

Source: USDA, BLS

Corn prices are around $4.75–$5.00/bu this fall, up from a long-term range of $2.00–$2.50/bu. We estimate that every $0.05/bu increase in corn prices has a 60 bps impact on cash margins for a generic ethanol plant, although changes in feedstock costs will vary according to the specifics of each plant design and hedging policy. Overall, we expect crush spreads (the price of ethanol less energy and net corn costs) to contract in 2009, leading to further margin pressure for the industry. Several years ago the DoE estimated that, longer term, the industry would need to switch to cellulosic feedstocks by the time production exceeds 12bn gal/year: this thesis could be tested as early as 2010 if all the proposed new corn ethanol plants are completed. We expect this to precipitate a shift in global corn trade flows, particularly to the benefit of Argentina and Brazil.

EXHIBIT 20: SUMMARY OF PRICING FOR ETHANOL, OIL AND BIOFUEL FEEDSTOCKS, 1998-2007

CAGR % from 2003-2007 Commodity Current 1998-2007 2003-2007 Peak Trough Ethanol ($/gal) $2.50 7.66% 16.28% -35% 126% WTI ($/bbl) $101 17.36% 31.29% -28% 129%Corn ($/bushel) $4.84 -0.34% 13.56% -31% 157% Palm Oil ($/t) $878 3.237% 17.174% -34% 150%Soy ($/bushel) $9.80 2.19% 9.45% -39% 92% Tallow ($/lb) $0.34 5.72% 14.53% -29% 147%

Source: Bloomberg

Industrial Biotech



Page 84 of 212

EXHIBIT 21: ETHANOL SPREAD VS. GASOLINE, AND ETHANOL CRUSH SPREAD, 1990-2008

Source: USDA, Bloomberg

EXHIBIT 22: OIL AND BIOFUEL FEEDSTOCK PRICES, INDEXED 1994=100. SYNCHRONIZED PEAK


Industrial Biotech


$(0.50)

$0

$0.50

$1.00

$1.50

$2.00

$2.50

$3.00

$3.50

Dec

-97

Jun-

98

Dec

-98

Jun-

99

Dec

-99

Jun-

00

Dec

-00

Jun-

01

Dec

-01

Jun-

02

Dec

-02

Jun-

03

Dec

-03

Jun-

04

Dec

-04

Jun-

05

Dec

-05

Jun-

06

Dec

-06

Jun-

07

Dec

-07

Jun-

08

Ethanol crush spread Ethanol spread vs. gasoline

$(0.50)

$0

$0.50

$1.00

$1.50

$2.00

$2.50

$3.00

$3.50

Dec

-97

Jun-

98

Dec

-98

Jun-

99

Dec

-99

Jun-

00

Dec

-00

Jun-

01

Dec

-01

Jun-

02

Dec

-02

Jun-

03

Dec

-03

Jun-

04

Dec

-04

Jun-

05

Dec

-05

Jun-

06

Dec

-06

Jun-

07

Dec

-07

Jun-

08

Ethanol crush spread Ethanol spread vs. gasoline

0

50

100

150

200

250

300

350

Jan-

94

Jan-

95

Jan-

96

Jan-

97

Jan-

98

Jan-

99

Jan-

00

Jan-

01

Jan-

02

Jan-

03

Jan-

04

Jan-

05

Jan-

06

Jan-

07

Jan-

08

CORN SOY TALLOW PALM OIL


Page 85 of 212

EXHIBIT 23: OIL AND BIOFUEL FEEDSTOCK PRICES, INDEXED YE04=100. PEAKING WITH CRUDE OIL


EXHIBIT 24: OIL AND BIOFUEL FEEDSTOCK MOMENTUM, 1998-2008: BIOBASED FEEDSTOCKS MORE SYNCHRONIZED SINCE 2003

EthanolTallow

Palm

SoyCorn

WTI

0%

5%

10%

15%

20%

25%

30%

35%

-2% 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20%

10 Year CAGR (%)

5Ye

arC

AG

R(%

)

Source: Bloomberg

Transportation

Ethanol is primarily distributed by truck, rail, or barge. Ethanol is hydrophilic, which can complicate using a pipelinebecause ethanol tends to separate from gasoline over time. As a result, it cannot be shipped in pipelines, andneeds to be blended at distribution terminals rather than at the refinery. This implies the need for a significantincrease in storage capacity over the next couple of years to accommodate the capacity additions that have alreadybeen announced. Ethanol’s miscibility issues also provide an incentive for dedicated transport fleets (less timespent cleaning containers), which should provide an opportunity for transportation companies to improve theirutilization rates. According to media reports, approximately 35%–40% of the new orders for railcars are for cars thatcan ship ethanol. Given the tight operating conditions in the railcar industry, however, we expect many of these newcars to be delivered in 2H07–2008. While rail capacity is a critical issue for ethanol producers, it is, in our view,immaterial for the railroads, as ethanol shipments represent only about 1% of total railcar loadings.

With this in mind, transportation costs are a significant factor for ethanol producers. Industry sources estimate thatin the US typical ethanol shipping costs run around $0.08/gal, vs. $0.02–$0.04/gal for tanker gasoline. Brazilianethanol imported into the United States faces higher shipping costs (estimated at $0.20–$0.25/gal), as well as a$0.54/gal import duty (to offset the $0.51/gal refiner’s subsidy), and, importantly, competition from consumer sugardemand. Many Brazilian ethanol plants are flexible and can switch from producing ethanol to producing refined

Industrial Biotech


0

50

100

150

200

250

300

350

Dec

-04

Feb-

05

Apr-

05

Jun-

05

Aug-

05

Oct

-05

Dec

-05

Feb-

06

Apr-

06

Jun-

06

Aug-

06

Oct

-06

Dec

-06

Feb-

07

Apr-

07

Jun-

07

Aug-

07

Oct

-07

Dec

-07

Feb-

08

Apr-

08

Jun-

08

Aug-

08

WTI CRUDE CORN SOY TALLOW PALM OIL


Page 86 of 212

sugar (13-15 lb of sugar/gallon of ethanol), so the limiting constraint is the arbitrage between ethanol export economics and refined sugar economics.

Yields, process & location essential to cost position

In the United States, roughly three-quarters of ethanol production is based on a process called “dry milling”. In this process, corn is ground into flour, then mixed with water and enzymes to make dextrose, which is fermented into alcohol. In a nutshell, this process transforms corn into roughly equal parts ethanol, carbon dioxide, and distiller dry grains, which are fed to cattle. As a result, the key levers that affect process economics are feedstock costs, the ratios between the outputs, and logistics.

As a rough rule of thumb, feedstock costs typically represent 65% or more of the total cost of production. Arbitraging corn costs from suppliers is essential. The wave of new ethanol capacity coming onstream will likely reduce the flow of corn between states within the United States, leading to regional price differentials in the cost of corn. Further contributing to opportunities for regional arbitrages will be disparities in natural gas costs (as much as 8% of the cost of producing ethanol), as well as the option of substituting coal for natural gas as the direct energy source for drying and steam production. As such, plants will likely see a fair amount of volatility in their position on the competitive cost curve depending on temporary shifts in regional supply/demand balances, as well as the relative value of co-products (particularly dried distiller grains). For example, locating a plant near a cattle farm so that the co-product distiller grains do not need to be dried can save as much as $0.04–$0.10/gal for a U.S. ethanol producer.

Some ethanol producers use an alternative process called “wet milling.” In this process, chemical solutions and mechanical processes (i.e., grinding) separate out corn oil, fiber, starch, and other components, much as a refinery separates different co-products from oil. Key co-products are gluten (a less attractive substitute for dry distillers grains) and yeast (of unreliable value). Dry mills require about half as much capital as wet mills, or $1.50–$2.00/gal of capacity, but wet mills can produce a broader range of co-products, including animal feed and corn syrup. In either case, we expect plant debottlenecking to contribute 1%–2% annual capacity creep.

Another way to improve ethanol economics is to genetically modify the feedstocks or the enzymes to enhance yields. This also has the benefit of reducing the incremental demand for nitrogen fertilizer, water, and real estate—essential to avoid adverse consumer sentiment on pollution and deforestation. DuPont and Monsanto are working on corn with higher starch content (a 2%–4% higher fermentation yield). Novozymes and Danisco have introduced enzymes they claim can decrease the capital intensity for some crops by 15%–20%. The jury is still out, however, to what extent ethanol producers will need to share with farmers the value created by modified crops.

Cellulosic ethanol: Flexible feedstocks and government incentives

Given mounting concerns that basing biofuel production on conventional food crops will lead to food prices rising in an arbitrage with oil on BTU content, governments and producers are allocating more resources to research into ways to convert other vegetable matter into biofuels. In essence, producers are also looking at ways to convert the carbohydrates in the cellulosic matter into sugars that can be fermented into ethanol. This means companies can process waste products such as corn stover, switchgrass, sawdust, and forest thinnings, as well as algae (from wastewater treatment plants).

Helping move this from theory to practice, the US DOE announced on February 28, 2007, an initiative to invest up to $385M into six biorefinery projects in 2007–2010. Industry investments in these projects will lift the total investment to more than $1.2bn. This program is expected to be supplemented by a loan-guarantee program for cellulosic ethanol, which we expect will broaden the field of competitors. Meanwhile, we expect other oil companies to follow BP’s strategy of embarking on large multi-year partnerships with cellulosic ethanol producers, such as its $90m 18-month partnership with Verenium.

Ethanol from cellulosic feedstocks is expected to eventually cost $0.60–$0.75/gal to produce, or $0.80–$1.00/gal per equivalent gallon of gasoline. This compares with current production economics north of $2/gal for most technologies. Taking into account debt financing (which could exceed $0.50/gal for one of Verenium’s first generation plants, for example), and integrated production technologies, Verenium estimates its first generation plants should be able to produce ethanol for roughly $1.90/gal, or more than $0.50/gal cheaper than the typical dry corn ethanol plant. The 2008 Farm bill introduced a $1.01/gal net blender’s credit for cellulosic ethanol. Also, the RFS mandate established a range for cellulosic ethanol pricing between $3/gal and $0.25 above gasoline prices (rising with inflation), depending on which is higher, in the event that the market lacks the capacity to exceed the mandate.

Industrial Biotech



Page 87 of 212

EXHIBIT 25: SUMMARY OF DOE FUNDING FOR CELLULOSIC BIOREFINERIES

Project Location Capacity (m gpy) Feedstock Funding Participants

Abengoa Missouri 11.4 700 tpd corn stover,

wheat straw, milo stubble, switchgrass

Up to $76M Abengoa, Antares, Taylor Engineering

ALICO Florida 13.9 770 tpd yard, wood and vegetative waste Up to $33M

Bioengineering Resources, Washington Group, GeoSyntec,

BG Katz and Emmaus Foundation

BlueFire Ethanol Southern California 19

700 tpd green waste and wood waste from

landfills Up to $40M Waste Management, JGC, MECS,

NAES and PetroDiamond

Poet Iowa 125

25% of production (31m gpy) will be cellulosic

ethanol, using 842tpd of corn fiber, cobs, and

stalks

Up to $80M DuPont, Novozymes, NREL

Iogen Idaho 18

700 tpd of what straw, barley straw, corn

stover, switchgrass, and rice straw

Up to $80M Iogen, Goldman Sachs, Shell

Range Fuels Georgia 401,200 tpd of wood residues and wood

energy crops Up to $76M

Merrick, PRAJ, Western Research Institute, Georgia Forestry

Commission, Yeomans Wood & Timber, BioConversion

Technology, Khosla Ventures, Gillis Ag & Timber, CH2MHill

Source: DOE

EXHIBIT 26: COMPOSITION OF GRAINS VS. CELLULOSICS (% OF TOTAL, +/- 5%)

Corn Grain Corn Stover Switchgrass Poplar Starch 72% 0% 0% 0% Cellulose/Hemicellulose 11% 69% 60% 72%Lignin 0% 16% 10% 21% Other Sugars 1% 4% 6% 3%Protein 9% 2% 5% 0% Oil/Extractives 5% 2% 13% 3%Ash 2% 7% 6% 1%

Source: NREL

Of particular interest, in our view, is the DuPont/NREL/Novozymes/Poet project, which aims to produce ethanol from both corn starch and from the corn stover—trying to improve plant economics by using more of the crop as a fuel feedstock. As a rough rule of thumb, carbon is just under 45% of the dry weight of corn, and only about a third of that is in the leaves. Part of the appeal of cellulosic ethanol, then, is the opportunity to access a greater portion of the carbon content in the crop, rather than wasting it. Complicating this, however, is the consideration that crop residues also serve to help fix carbon in topsoil. As a result, farmers have raised the concern that using corn stover for cellulosic ethanol will end up “strip mining the topsoil,” leading to unsustainable longer-term process economics.

Shifting feedstocks could reduce the environmental impact. Switchgrass, for example, requires roughly 25% of the fungicide, 10% the insecticides, and half the fertilizer of corn, wheat, or soybean crops. Using switchgrass can also reduce the amount of nitrogen run-off by as much as 90% (vs. corn) or 50% (vs. soybeans). The NRDC, meanwhile, has further estimated that switchgrass can reduce soil erosion by north of 95%. Another appeal of using cellulosic feedstocks, in our view, is the opportunity to leverage the existing infrastructure of the North American paper industry.

EXHIBIT 27: STEPS IN BIOMASS CONVERSION

Thermochemical treatment of raw biomass

Render complex polymers (cellulose, hemicellulose, lignin) more accessible to enzymatic breakdown

Enzyme applications (cellulases and hemicellulases)

Break down (hydrolyze) plant cell-wall constituents (polysaccharides) into simple sugars

Fermentation Bacteria and yeast convert sugars to ethanol Source: DOE

Industrial Biotech



Page 88 of 212

Other (increasingly) important factors that will drive cellulosic ethanol economics include the cost of procuring biomass (collecting and storing the material), pre-treating the biomass (part of the overall cost of the process, also affecting capital intensity), optimizing plant designs (to reduce labor costs and capital intensity), opportunities to modify yeast (as a substitute for new enzymes), using alternative, more genetically-flexible microbes (move to single-step biotransformations) and government subsidies (rising globally as governments seek to hasten the diversification of energy sources). Improving yields (gal/acre) is critical, in our view, as higher yields should translate into economies of scale by reducing logistics costs for biomass harvesting.

EXHIBIT 28: THEORETICAL YIELD OF ALTERNATE FEEDSTOCKS (ETHANOL GAL/TON OF FEEDSTOCK)

Corn Grain 124.4 Corn Stover 113.0 Rice Straw 109.9 Cotton Gin Trash 56.8 Forest Thinnings 81.5 Hardwood Sawdust 100.8 Bagasse 111.5 Mixed Paper 116.2

Source: DOE

Industrial Biotech



Page 89 of 212

Biodiesel: Focus on Feedstock Arbitrage

Biodiesel is a transportation fuel derived from animal fats, oils and greases, as well as vegetable oils. More specifically, biodiesel is an ester of fatty acids, derived from animal fats, waste oils and fats, fatty acids from waste treatment facilities, and vegetable oils such as rapeseed and soybean. The industry has received significant government support, including a federal mandate for the use of 1bn gpy of biodiesel by 2010.

EXHIBIT 29: GOVERNMENT INCENTIVES: TAX CREDITS, MANDATES AND INVESTMENT CREDITS

Tax Use Investment

Federal $1/gal blender's credit for virgin feedstock, $0.50/gal for recycled feedstock. Due to expire Dec. 31, 2008, with an extension in the Farm Bill.

2007 Energy bill set a mandate of 1bn gallons of biodiesel by 2012.

30% tax credit for installing B20-or better equipment at refueling stations

States Credits and incentives in 14 states. Several states have mandated biodiesel use in public vehicles.

Sundry credits for producing fuel, converting vehicles, or installing refueling stations

Source: Company information

Biodiesel is produced via transesterification, a relatively simple refining process that converts roughly 98% of vegetable oil to biodiesel. Methanol and catalysts convert fat or oil into glycerin and methyl esters (the biodiesel). The glycerin is sold to producers of soap, cosmetics or, at higher purity levels, pharmaceuticals. An alternative process would use ethanol rather than methanol; this has not been commercialized on a significant scale. Net-net, the standard process transforms feedstock (typically soy oil in the US, 87% of inputs), catalysts (1%), and alcohol (typically methanol, 12%) into biodiesel (methyl ester, 86%), glycerine (9%), alcohol (4%), and fertilizer (1%).

Biodiesel can be mixed with regular diesel at any proportion: it has slightly less energy content (8% lower), but higher fuel density and a higher cetane number. When mixed with fossil diesel, it is referred to by the percentage of biodiesel in the blend. For example, B5 has 5% biodiesel, B10 10%, and so forth. From an environmental standpoint, the appeal of biodiesel stems from the fact that it is biodegradable, non-toxic, generates roughly 78% less CO2 than conventional diesel, reduces particulate emissions by 55% and other emissions by 60%–90%—with the notable exception of NOx emissions, which are roughly 10% higher but are easily filtered. Importantly, biodiesel blends can be used in the existing liquid transportation fuel distribution infrastructure, without the retrofit investments that accompany increased use of ethanol in gasoline.

EXHIBIT 30: COMPARISON OF CONVENTIONAL TRANSESTERIFICATION PROCESSES

Input Methanol Ethanol Property Methyl Ester Ethyl Ester Conversion factor (oil to biodiesel) 97.50% 94.30% Total glycerin (%) 0.86% 1.40% Viscosity - 7.2% higher % power vs. diesel < 2.5% < 4%

Source: BMVEL

EXHIBIT 31: TYPICAL BIODIESEL MASS BALANCE

Inputs % Outputs %Vegetable Oil 87% Biodiesel 86% Alcohol 12% Glycerine 9%Catalyst 1% Alcohol 4%

Fertilizer 1%Source: NBB

Industrial Biotech



Page 90 of 212

EXHIBIT 32: BIODIESEL ENERGY DENSITY COMPARED TO NO. 2 DIESEL AND OTHER ENERGY SOURCES (BTU/GAL)

1.0 3.4 9.8

76.385.0

116.1 118.3129.3 129.5

0

20

40

60

80

100

120

140

Nat

ural

Gas

(BTU

/scf

)

Ele

ctric

ity(B

TU/k

wh)

Coa

l(B

TU/lb

)

Eth

anol

LPG

Gas

olin

e

Bio

dies

el

B2

blen

d

No.

2di

esel

1.0 3.4 9.8

76.385.0

116.1 118.3129.3 129.5

0

20

40

60

80

100

120

140

Nat

ural

Gas

(BTU

/scf

)

Ele

ctric

ity(B

TU/k

wh)

Coa

l(B

TU/lb

)

Eth

anol

LPG

Gas

olin

e

Bio

dies

el

B2

blen

d

No.

2di

esel

Source: USDA, NBB

Large addressable market, constrained by feedstock supply

The addressable markets for biodiesel include the 63bn gal/year market for diesel in the United States, the 85bn gal/year market for diesel in Europe, and a roughly 14bn gal/year market in the U.S. for home heating fuel. These markets are expected to grow roughly 1%–2% a year.

Faced with these perceived opportunities for substitution, government mandates should continue to drive capacity additions in the biodiesel industry. Indeed, global biodiesel capacity is expected to reach roughly 6bn gal/year by 2009, with Europe representing almost two-thirds of that. Utilization rates, however, are relatively low due to technological challenges and pressure from rising feedstock costs. Earlier this year, the National Biodiesel Board estimated that the 171 plants in the United States had 2.24bn gal/year of installed capacity, with another 1.1bn gal/year of capacity planned for the next couple of years. This would bring U.S. biodiesel capacity by 2010 to roughly 3x the proposed federal mandate for 2013.

In the U.S., domestic production is expected to reach 450m gal/year in 2007 (0.7% of diesel demand), up from only 25m gal/year in 2004 and 2m gal/year in 2000. To put these figures in context, were biodiesel to reach 2% of total U.S. diesel consumption, U.S. demand alone would be approximately 1bn gal/year.

EXHIBIT 33: U.S. BIODIESEL PRODUCTION (M GAL/YEAR), VS. 60BN GAL/YEAR OF DIESEL CONSUMPTION

0.5 2 5 15 20 2575

250

450

050

100150200250300350400450500

1999 2000 2001 2002 2003 2004 2005 2006 2007

0.5 2 5 15 20 2575

250

450

050

100150200250300350400450500

1999 2000 2001 2002 2003 2004 2005 2006 2007


Rather than end market growth, however, the fundamental constraint is feedstock supply: in the 2008 planting season, biodiesel is estimated to represent an estimated 20% of U.S. soy oil demand and 60% of European rapeseed oil demand. Another way to frame the constraint is to consider that using all of the global supply of vegetable oils and fats (roughly 80m tpy of feedstock) would only support 22–24bn gal/year of biodiesel production, or 13%–15% of the ostensible addressable market in the U.S. and Europe. Similarly, using all of the U.S. supply of vegetable oils and fats would support 4bn gal/year of biodiesel production.

Industrial Biotech



Page 91 of 212

EXHIBIT 34: EU AND U.S. BIODIESEL PRODUCTION BY FEEDSTOCK (M GAL/YEAR), OCT.-SEP. PRODUCTION YEARS

2006 2007E 2008E 2009E 2010E 2011E 2012E 2013E 2014E 2015E 2016E U.S. (2007 outlook) 385 541 569 578 565 551 534 511 491 472 449 Proposed RFS 500 650 800 1,000 1,000 1,000 1,000 1,000

From Soybean Oil 331 450 467 469 450 430 410 386 367 348 324 From Canola Oil 30 62 71 76 80 85 85 84 81 79 78 From Other Fats

& Oils 24 29 31 33 35 37 39 41 43 45 47 Gap vs. RFS 78 -85 -249 -466 -489 -509 -528 -551 % of total

From Soybean Oil 86% 83% 82% 81% 80% 78% 77% 76% 75% 74% 72% From Canola Oil 8% 12% 13% 13% 14% 15% 16% 16% 17% 17% 17% From Other Fats 6% 5% 5% 6% 6% 7% 7% 8% 9% 9% 10%

EU 5,504 6,157 6,432 6,526 6,558 6,546 6,639 6,797 6,977 7,161 7,343 Rapeseed Oil 4,675 5,499 5,725 5,820 5,834 5,801 5,871 6,012 6,180 6,362 6,542 Soybean Oil 1,395 1,245 1,264 1,233 1,217 1,201 1,215 1,220 1,236 1,242 1,246 Sunflower Oil 95 152 216 256 294 330 349 381 399 417 436 % of total

Rapeseed Oil 85% 89% 89% 89% 89% 89% 88% 88% 89% 89% 89% Soybean Oil 25% 20% 20% 19% 19% 18% 18% 18% 18% 17% 17% Sunflower Oil 2% 2% 3% 4% 4% 5% 5% 6% 6% 6% 6%

Source: FAPRI, NBB

Globally, if the U.S. maintains its target of 1bn gal/year of biodiesel by 2012, combined EU and U.S. production appears slated to increase to 7.6bn gal/year in 2012 from 5.9bn gal/year in 2006. With 57 new biodiesel plants under construction in the U.S. alone, however, we believe these estimates could prove conservative, particularly if oil prices remain north of $80/bbl.

There is a constraint, however, to government ambitions to offset the 345bn gal/year demand for petro-diesel (62bn gal/year in the U.S. alone). Even if the most popular biodiesel feedstocks—soybean oil, rapeseed oil and palm oil—were switched entirely to biodiesel production, global biodiesel production would only reach an estimated 34bn gal/year, or 10% of current global diesel demand. This would imply, of course, no vegetable oils available for food, feed or oleochemical applications. Clearly, any further push to use these vegetable oils as feedstocks would likely entail a further (potentially sharp) increase in prices due to the competing needs of different stakeholders.

EXHIBIT 35: GLOBAL OILSEED OIL PRODUCTION (‘000 T) AND BIODIESEL EQUIVALENT (BN GALLONS)

2007E 2008E 2009E 2010E 2011E 2012E 2013E 2014E 2015E 2016E 2017E Soybean Production (m t) 35.67 36.42 37.14 38.21 39.32 40.43 41.56 42.69 43.84 45 46.18 Biodiesel equivalent (bn gal) 9.826 10.03 10.23 10.53 10.83 11.14 11.45 11.76 12.08 12.4 12.72

Rapeseed oil Production (m t) 17.78 18.76 19.18 19.58 19.89 20.16 20.47 20.81 21.16 21.53 21.9 Biodiesel equivalent (bn gal) 4.898 5.169 5.283 5.395 5.481 5.555 5.639 5.733 5.831 5.931 6.032

Palm oil Production (m t) 38.97 40.57 42.33 44.05 45.76 47.45 49.15 50.88 52.64 54.44 56.26 Biodiesel equivalent (bn gal) 10.74 11.18 11.66 12.14 12.61 13.07 13.54 14.02 14.5 15 15.5

Total biodiesel equivalent 25.46 26.38 27.18 28.06 28.92 29.76 30.63 31.51 32.41 33.33 34.26

Source: FAPRI, NBB, Jefferies & Company, Inc. estimates

Multiple feedstocks create arbitrage opportunities

One appeal of biodiesel is feedstock flexibility. The proposed RFS mandate of 1bn gallons of biodiesel production by 2013 implies demand for 7.5–8.0bn gallons of feedstock. This would represent almost 35% of all the vegetable

Industrial Biotech



Page 92 of 212

oil production in the U.S. at current rates. The only way to alleviate the implied upward pressure on feedstock prices is to source alternate materials.

The available biodiesel technology can process fats, waste oils, or vegetable oils, which can be esterified into methyl or ethyl groups. Vegetable oils are extracted by pressing or distilling soybeans, cotton, sunflower seeds, rapeseed, corn, palm seeds, coconuts, and other oil-rich fruits. Acceptable fats, on the other hand, include tallow, pork fat, chicken fat, yellow grease, and cooking oil.

Naturally, which feedstock will be favored depends on local economics, particularly the availability of storage infrastructure and the cost of logistics (production centers are often removed from consumer centers). In the United States, soybeans account for more than 90% of biodiesel production, as they have relatively low levels of free fatty acids (FFAs) and most plants used technology that saw sharp yield losses when attempting to process feedstocks with higher FFA levels. Not surprisingly, these plants have seen process economics deteriorate due to the sharp rise in row crop prices over the past couple of years.

Overall, we expect producers with a flexible approach and access to large, well established global markets such as palm oil to face less feedstock price volatility over time than producers tied to a single crop. Waste feedstocks, such as inedible tallow and yellow grease, often have better economics, but supply is more limited. Indeed, even if the U.S. biodiesel industry can convert all the available grease, soybean oil, and tallow into biodiesel, we estimate it would only be large enough to displace 5% of current diesel demand (vs. 0.1% last year). Moreover, regional shifts in meat production (perhaps prompted by higher grain prices?) could have an impact on the availability of related waste products.

EXHIBIT 36: Y2007 BIODIESEL CAPACITY (2.2BN GAL) BASED ON PRIMARY FEEDSTOCK

Soy38%

Other9%

MultiFeedstock

51%

RecycledCooking Oil

0%

YellowGrease,

Animal Fat2%

Source: NBB

EXHIBIT 37: PROPOSED NEW U.S. BIODIESEL CAPACITY (M GAL/YEAR OF CAPACITY) COMING ONSTREAM IN 2008

1Q08 2Q08 3Q08 4Q08 2008Total 402.2 227 191.5 243 1,063.7Includes Soy-based 95 105 45.5 60 305.5

Source: NBB

Approximately half of U.S. biodiesel installed capacity, and roughly 70% of the capacity scheduled to come online in2008, claims to be able to process multiple feedstocks. We believe, however, that most of these plants still suffersignificant yield losses when attempting to process higher-FFA feedstock (north of 8%).

Key U.S. feedstocks that could support 4.0–4.3bn gal/year of biodiesel production include:

• Vegetable oils: 23.2bn lbs in the U.S., including oils derived from soybean (95% of US biodiesel productionin practice), rapeseed (commonly used in Europe), palm oil, cottonseed, groundnut, sunflower, corn(DDGs), coconut, olive, castor, sesame or linseed oil.

• Animal fats: 6.2bn lbs, encompassing both edible fats such as tallow and white grease with free fatty acidlevels up to 0.8%, and inedible fats with FFA levels above 10%.

• Recycled cooking oils and greases: 2.6bn lbs. Yellow grease typically has FFA levels less than 15%, whilebrown grease typically has FFA levels above 15%. These are usually the cheapest feedstocks, due to the

Industrial Biotech



Page 93 of 212

lack of direct competition with human consumption requirements. We note, however, that even yellow grease has seen a 75% increase in prices from historical levels, to $0.20–$0.25/lb, almost comparable to the 90% increase in tallow prices (to $0.30/lb) and the 150% increase in soy oil prices (to $0.55/lb).

EXHIBIT 38: COMPARISON OF SOURCES OF VEGETABLE OIL/BIODIESEL FEEDSTOCK

Plant Source Oil Content (%) Months of harvesting Yield (t/ha) Algae Organism 50 nmf 35-125 Palm Pulp 26 12 3.0-6.0 Sunflower Seed 38-48 3 0.5-1.5 Rapeseed Seed 40-48 3 0.5-0.9 Castor Seed 43-45 3 0.5-1.0 Peanuts Seed 40-50 3 0.6-0.8 Soybeans Seed 17 3 0.2-0.6

Source: BMVEL

By using high-fatty acid waste products and animal fats, it should be possible for some producers to maintain a lower position on the cost curve than the producers using soybean oil. Since the turn of the century, and again using only generic process economics and current (elevated) ancillary consumable costs, relatively expensive feedstocks have been profitable after taking into account biodiesel’s lower energy density vs. conventional diesel, as well as SG&A and depreciation costs. Biodiesel production based on soybean oil, in contrast, has tended to generate favorable crush margins while failing to deliver operating profits after taking into account depreciation—much as one would expect for marginal supply.

Setting industry standards EXHIBIT 39: ASTM 6751 BIODIESEL SPECIFICATIONS

Criterion ASTM6751 Nova BenefitsFree glycerin, mass % (max) 0.02% < 0.02% Lower free glycerin results in reduced engine depositsTotal glycerin, mass % (max) 0.24% < 0.05% Fewer unconverted glycerides reduces particulate emisions and engine depositsFlash point, deg C (min) 130 C > 170 C High purity biodiesel has a higher flash pointWater & sediment, vol % (max) 0.05% < 0.01% Water and sediment promote biological growth and filter pluggingKinematic Viscosity, cSt @40 deg C 1.9 - 6.0 3.9 - 4.5 High purity biodiesel has a tight viscosity rangeSulfated Ash, mass % (max) 0.020% < 0.005% Low sulfer biodiesel reduces sulfated ash formation, particulate emissionsTotal sulfur, mass % (max) 15 ppm < 15 ppm Low sulfur content meets S15 designation, same as ULSDCopper corrossion (max) No. 3 No. 1a High purity biodiesel has a very low corrosivityCetane Number (min) 47 > 53 Dependent on feedstock and additive treatmentsCloud point, deg C report report Dependent on feedstock and additive treatmentsCarbon residue, mass % (max) 0.05% < 0.02% High purity biodiesel burns more completely, reduces particulate emissionsAcid number, mg KOH/g (max) 0.5 < 0.5 Low acid number increases long term storage stabilityPhosphorus, mass % (max) 10 ppm < 5 ppm Low phosphorus concentrations reduce particulate emissions and engine depositsDistillation 90% recovery (max) 360 C < 360 C Lower recovery temperature is indicative of higher purity biodieselNa+ and K+, mass % (max) 5 ppm < 3 ppm Low metals concentrations reduces particulate emissions and injector foulingCa2+ and Mg2+, mass % (max) 5 ppm < 3 ppm Low metals concentrations reduces particulate emissions and injector foulingOxidative stability (min) 3 hrs > 3 hrs Improves long term storage stabilityCold stock filtration (max) * 360 sec < 200 sec Improved cold weather performanceWater, Karl Fischer titration (max) * 500 ppm 100 ppm Low water concentrations prevent biological growth while biodiesel is in storageSediment, mass % (max) * 24 ppm 5 ppm Low sediment concentrations reduce filter plugging issues

Specifications

* Additional specification anticipated in upcoming revisions of D6751Bold = BQ-9000 critical specification testing once production process under control

Source: National Biodiesel Board (NBB) and company information

ASTM standards for biodiesel are expected to tighten at the June ASTM meeting. With biodiesel blends already available in more than 1,250 retail filling stations (vs. 1,200 currently offering E85 gasoline), more stringent standards are viewed as a key step for broader market acceptance. The new standards would address several key concerns:

• Specification: The biodiesel portion of B6 to B20 will need to meet the standard for B100 before blending, and the finished blend must satisfy the specifications for No. 1 or No. 2 diesel.

• Cloudpoint and related filter clogging: Adjustments to the B100 standard should reduce the performance issues that have plagued the industry to date.

• New applications: Blends for diesel fuel and home heating oil will be set for up to B5 (5% biodiesel) content.

Industrial Biotech



Page 94 of 212


Environmental controversies

From most perspectives, biodiesel represents an attractive alternative to conventional diesel, either as a direct substitute or as a way to establish “peaking” capacity—particularly in the United States where building new refinery capacity remains controversial. There remain, however, some environmental concerns that could undermine government support for biodiesel mandates. The first concern is that untreated B100 (100% biodiesel) emits roughly 10% more NOx emissions than regular diesel. This should be easily addressed with fuel additives and catalytic converters, but has proved to be a popular talking point. The second, more controversial concern is that government mandates to use biodiesel, by supporting feedstock prices, encourage deforestation in developing countries. In the near term, this concern is being addressed by governments exploring mandates that specify feedstock sourcing in more detail, and by companies that establish “green credits” vetting the life-cycle impact analyses of different feedstocks. In the medium-term, in contrast, we expect these concerns to be addressed by shifting to alternative feedstocks, perhaps algae, that represent a more efficient use of waste streams rather than an incentive to plant traditional row crops on more acres.

Cloudpoint

In most respects, biodiesel represents an attractive alternative as a transportation fuel: similar energy density when compared with regular diesel; favorable lubricity; reduced greenhouse gas emissions; and so forth. A significant concern, however, is biodiesel’s cloudpoint. Specifically, biodiesel, on its own, can begin to congeal around 14–68 degrees F, necessitating the use of heated storage vessels and blending tanks. The cloudpoint is related to the mix of esters in the biodiesel, and consequently on the feedstock, with a feedstock like tallow tending to congeal at the higher end of the range and certain seed oils, such as canola, congealing at the low point. This issue is less noticeable when biodiesel is blended with regular diesel: at B5 or less, the cloudpoints are basically the same. While this issue appears addressable through both additives and seasonal blending regulations (i.e., requiring lower blends in the winter), it could prove a sticking point with respect to acceptance by consumers and distributors.

Water contamination

When biodiesel contains unreacted monoglycerides and triglycerides, they can attract water into the biodiesel, which is otherwise immiscible. This contamination can lead to the fuel congealing at a warmer temperature (as the water cools and sets into ice crystals), as well as supporting corrosion, microbial contamination, and engine component deterioration. Nova’s proprietary process, unlike many of its competitors, is not a water-based process, but the reputation of the industry could suffer due to competitors’ failure to maintain product quality standards.

Industrial Biotech


Page 95 of 212

Laur

ence

Ale

xand

er, C

FA

, lal

exan

der@

Jeffe

ries.

com

, (21

2) 2

84-2

553

EXH

IBIT

40:A

GR

ICU

LTU

RAL

BIO

REF

INER

YPA

THW

AYS,

INC

LUD

ING

DO

EPA

RTN

ERS

Resid

ualS

tarc

hCo

nver

sion

Com

plete

syste

msl

evel

Corn

Fibe

rCon

vers

ionde

mon

strat

ionan

dva

lidat

iono f

Grain

New

Frac

tiona

tion

Proc

ess

allte

chno

logies

toim

prov

ePr

oduc

tsfro

mC5

/C8

Suga

rsco

rnwe

tmill

facil

itiesu

sing

Prod

ucts

from

Oils

corn

grain

feed

stock

Com

plete

syste

msl

evel

Resid

ualS

tarc

hCo

nver

sion

dem

onstr

ation

and

valid

ation

ofOt

her

DDG

Conv

ersio

nall

tech

nolog

iesto

impr

ove

Grain

sNe

wFr

actio

natio

nPr

oces

sco

rndr

ymill

facil

itiesu

sing

Prod

ucts

from

C5/C

8Su

gars

corn

(and

othe

r)gr

ain

feed

stock

Com

plete

syste

msl

evel

Oil

dem

onstr

ation

and

valid

ation

ofCr

ops

Prod

ucts

from

Oils

allte

chno

logies

toim

prov

ena

tura

loilp

roce

ssin

gfa

cilitie

sus

ingoil

crop

feed

stock

Biom

assS

ugar

Prod

uctio

nCo

rnSt

ove r

Prod

ucts

from

C5/C

8Su

gars

Com

plete

syste

msl

evel

Small

Grai

nPr

oduc

tsfro

mLig

ninde

mon

strat

ionan

dva

lidat

ionof

Stra

wBi

omas

sGas

ificat

ionall

tech

nolog

iesto

impr

ove

Rice

Stra

wPr

oduc

tsfro

mSy

nthe

sisGa

spr

oces

sing

facil

itiesu

sing

New

Frac

tiona

tion

Proc

esse

sag

ricult

ure

resid

uefe

edsto

cks

Prod

ucts

from

New

Proc

essI

nter

med

iates

Biom

assS

ugar

Prod

uctio

nGr

asse

sPr

oduc

tsfro

mC5

/C8

Suga

rsCo

mple

tesy

stem

slev

elW

oody

Prod

ucts

from

Lignin

dem

onstr

ation

and

valid

ation

o fCr

ops

Biom

assG

asific

ation

allte

chno

logies

toim

prov

ePr

oduc

tsfro

mSy

nthe

sisGa

spr

oces

sing

facil

itiesu

sing

New

Frac

tiona

tion

Proc

esse

spe

renn

ialcr

ops.

Prod

ucts

from

New

Proc

essI

nter

med

iates

New

Cons

olida

ted

Proc

esse

s

ConversionPathwayOptions

UnderConsideration(eachhas

aBmilestone-costtarget)

Program"A"

Milestones

IndustrialPartners

Agric

ultu

reRe

sidu

es

ADM

Aben

goa

Dupo

nt

ADM

NCGA

Carg

ill

Pere

nnia

lCro

ps

BiorefineryPathway

Feedstocks

Wet

and

Dry

Mill

OilC

rops

Natu

reW

orks

,LLC

Char

iton

Valle

yGr

eatP

lains

Insti

tute

Univ.

ofTe

nnes

see

Aven

tine

Natu

reW

orks

,LLC

Carg

ill

Aben

goa

Broin

Dupo

nt

Sou

rce:

DoE

Ind

ust

rial

Bio

tech

Ple

ase

see

impo

rtan

t dis

clos

ure

info

rmat

ion

on p

ages

208

- 2

10 o

f thi

s re

port

.P

age

96 o

f 212

Laur

ence

Ale

xand

er, C

FA

, lal

exan

der@

Jeffe

ries.

com

, (21

2) 2

84-2

553

EXH

IBIT

41:F

OR

EST

BIO

REF

INER

YPA

THW

AYS,

INC

LUD

ING

DO

EPA

RTN

ERS

Woo

dNe

wFr

actio

natio

nPro

cess

esfor

Comp

letes

ystem

slev

elMi

llhe

mice

llulou

srem

oval

demo

nstra

tiona

ndva

lidati

onof

Was

tesPr

oduc

tsfro

mC5

/C6S

ugar

sall

techn

ologie

stoi

mpro

veBl

ackL

iquor

Gasif

icatio

nco

rnwe

tmill

facilit

iesus

ingPr

oduc

tsfro

mBL

GSy

n.Ga

sco

rngr

ainfee

dstoc

k

Comp

letes

ystem

slev

elW

ood

demo

nstra

tiona

ndva

lidati

onof

Mill

Pyro

lysis

OilU

pgra

ding

alltec

hnolo

giest

oimp

rove

Was

tesco

rndr

ymill

facilit

iesus

ingco

rngr

ainfee

dstoc

k.

Logg

ingBi

omas

sSug

arPr

oduc

tion

resid

ues

Prod

ucts

from

C5/C

6Sug

ars

Comp

letes

ystem

slev

elFu

elPr

oduc

tsfro

mLig

ninde

mons

tratio

nand

valid

ation

o ftre

atmen

tsBi

omas

sGas

ificati

onall

techn

ologie

stoi

mpro

vePr

oduc

tsfro

mSy

nthes

isGa

sna

tural

oilpr

oces

singf

acilit

iesNe

wFr

actio

natio

nPro

cess

esus

ingoil

cropf

eeds

tock

Prod

ucts

from

New

Proc

essI

nterm

ediat

es

Biom

assS

ugar

Prod

uctio

nGr

asse

sPr

oduc

tsfro

mC5

/C6S

ugar

sCo

mplet

esys

temsl

evel

Woo

dyPr

oduc

tsfro

mLig

ninde

mons

tratio

nand

valid

ation

o fCr

ops

Biom

assG

asific

ation

alltec

hnolo

giest

oimp

rove

Prod

ucts

from

Synth

esis

Gas

proc

essin

gfac

ilities

using

New

Frac

tiona

tionP

roce

sses

agric

ultur

alre

sidue

feeds

tocks

Prod

ucts

from

New

Proc

essI

nterm

ediat

es

None

None

Non-

Fore

stW

ood

Was

tes

BiorefineryPathway

Feedstocks

Pulp

and

Pape

rMills

Fore

stPr

oduc

tMills

ConversionPathwayOptions

UnderConsideration(eachhas

aBmilestone-costtarget)

Program"A"

Milestones

IndustrialPartners

Fore

stRe

sidue

sNo

ne

Geor

giaPa

cific

Agen

da20

20

Sou

rce:

DoE

Ind

ust

rial

Bio

tech

Pag

e 97

of 2

12P

leas

e se

e im

port

ant d

iscl

osur

e in

form

atio

n on

pag

es 2

08 -

210

of t

his

repo

rt.




EventCapital investments by ADM, BASF, Braskem, Cargill, DSM, DuPontand Teijin Chemicals have helped draw attention to bioplastics. Thisreport reviews the emerging industry, including key players, growthdrivers, challenges and likely investment themes.

Key Points• Cost and brand considerations drive adoption. Surveys

conducted in mid-2007 showed that 72% of U.S. respondents didnot know that plastics were derived from oil or natural gas, and 40%thought that plastics biodegrade. Consumer demand created byfocused branding efforts, opportunities to reduce energy and wastehandling costs, and regulatory drivers should drive 15%-plus annualgrowth in the estimated 120,000-135,000tpy market forbiodegradable plastics (vs. 190,000-215,000tpy based onrenewable raw materials).

• Several ways to optimize yields: In classic fermentation,companies modify organisms so they are more tolerant of theirmonomer by-products (make the yeast a happy alcoholic). In the"foie gras" method, companies modify metabolic feedback loops soorganisms store excess food as polymers, effectively compressingseveral chemical steps into a single biological pathway. The former,in our view, is better suited for producing bio-based chemicals, thelatter for bio-based polymers (lower capital costs). Somecompanies are already exploring ways to produce biodegradableplastics via non-biological catalytic routes, which could prove easierto scale-up to industrial capacity. With the field attracting newentrants rapidly, integrated R&D and production platforms couldprove essential to maintaining a competitive advantage.

• Bio-based chemicals could compete for feedstocks. Enzymesand cellulose-based products have been the highest-profilebio-based chemicals. We expect the industrial biotech industry toexplore alternatives to petrochemicals that are more renewable(good for branding), biodegradable (reducing waste costs), andeconomical (given stubbornly high oil prices). Sugarcane,switchgrass, palm oil, soybeans, and waste cellulose could offerbetter economics than corn starch in some contexts.

• Alternatives to new polymers? Whereas companies likeNatureworks and Metabolix have focused on new bio-basedpolymers, the increasing interest in bio-based plastics has led totwo further developments: the conversion of ethanol to polyethyleneand PVC (pioneered in Brazil in the 1980s); and dedicatedmonomer production (whether bio-PDO or lactide) that othercompanies can use to make bio-based materials on proprietaryformulations. CMAI has estimated that to produce 1m tpy ofethylene (1% of global demand) from corn would take 2% of the2006 corn harvest and 14% of 2006 ethanol production.

October 6, 2008


Clean TechnologyBioplastics Primer

Investment SummaryBio-based and biodegradable plastics are emerging as an attractiveniche product that downstream processors can use to addressfeedstock pressures while improving both their brands and thescarcity value of their operations. Metabolix (MBLX, $10.31, Hold,$10.50 PT) provides a pure-play on this theme.


Robin Campbell, Ph.D., Equity Analyst44 (0) 20 7029 8678, [email protected]



Overview

Nature does not eat plastic, yet. Microbes that have had thousands of years to learn to eat cellulose, lignin, and even oil, are about as comfortable eating plastic as vegetarians are eating rock. While certain fungi appear able to digest phenolic resins, PCBs and even polystyrene, the most reasonable case for most of the roughly 1bn tons of plastic produced since 1950 globally is that it will still be around in several centuries’ time. While large units of plastic, if discarded rather than recycled, might be expected to settle as sediment, eventually creating impromptu landfills, smaller units, such as those used in body scrubs and hand cleaners, are small enough to float on the ocean currents for thousands of years. Even “green” plastics that are intended to photodegrade or, by blending petrochemicals with starch-based polymers, biodegrade, will still leave behind small fragments of polymer. Moreover, these fragments appear to attract and absorb other wastes, such as PCBs and DDTs, sharply increasing their concentration in seawater.

Until regulators or waste treatment facilities tie the waste products back to consumers or producers as a cost, these considerations, while concerning for some, are largely immaterial as economic factors. For companies that can produce polymers that biodegrade in natural environments, “natural plastics” as it were, and for investors, what matters is that there are applications for biodegradable plastics that can deliver economic value in the near term. Being able to leave erosion control stakes in the ground, for example, saves the cost of hiring staff to pick them up again. For other applications, the economics are driven by consumer sentiment, as certain consumers prefer to think that garbage will not be their most permanent legacy. Sentiment, however, can prove to be a thin reed when household budgets come under pressure.

EXHIBIT 1: DEVELOPMENT STAGE OF DIFFERENT BIOPOLYMERS

Research Development Pilot Plant Commercial

Large Scale

Industrial Scale

2007 Bio-Ethylene (Bio-PE) xBiopolyamides (Bio-PA) x xBiopolyurethanes (Bio-PUR) x xCellulose Acetates (CA) xDegradable Polyesters xPolybutyleneterephthalates (PBT) xPolybutylsuccinates (PBS) xPolycaprolactones (PCL) xPolyhydroxyalkanoates, butyrates, etc. (PHAs, PHBs)

x x

Polylactides (PLA) xPolyvinylalcohols (PVOH) xStarch Blends x2010E Bio-PE x xBio-PA x xBio-PUR x xCA xDegradable Polyesters xPBT x xPBS xPCL xPHA/PHB x xPLA xPVOH xStarch Blends x

Source: Jefferies & Company, Inc. estimates, Bioplastics magazine

Industrial Biotech



Page 100 of 212

While the biodegradable plastic industry is still in its infancy, in our view, polymer producers and processors already recognize the importance of securing a diverse range of feedstocks. While much of this report is focused on biodegradable plastics rather than bio-based ones, the latter opportunity is likely to eventually emerge as the larger product category (albeit also more hotly contested) as traditional petrochemical producers look for ways to alleviate their risk profile in terms of dependence on oil and natural gas as feedstocks.

1. How large is the market?

Supply/demand: We estimate global demand for biodegradable plastics at roughly 120,000–135,000 tpy (roughly 65% in Europe, 22% in North America, 10%–12% in Japan), up from 85,000–100,000tpy in 2005. We expect demand to grow 10%–15% through the end of the decade. We estimate the industry is operating at roughly 35% of nameplate capacity, in part because some facilities have effective production constraints well below their nameplate figures.

EXHIBIT 2: GLOBAL BIOPLASTICS DEMAND BY APPLICATION (%), 2005

Compost Bags45%

Loose Fill27%

Packaging Film14%

Other14%

Source: Jefferies & Company, Inc. estimates, PlasticsTechnology Online

Robust growth in supply: The size of the market is obscured by inconsistencies in the available data (e.g., bio-based vs. degradable plastics), but we estimate bio-based (but not biodegradable) commodity chemicals could represent a market 2–4x as large as the market for biodegradable plastics. Given Natureworks’ recent announcement that its effective capacity is roughly half of the name-plate capacity reported since 2003, we have lowered our estimate for the aggregate installed capacity for biobased plastics to 365,000 tpy in 2007 and 370,000 tpy in 2008 (at least two-thirds biodegradable). Based on announced capacity additions, we estimate total capacity should increase 120% by the end of 2009 and 315% by 2013. This could prove aggressive, however, as the recent trend in the industry has been towards project delays (e.g., Metabolix’s new facility postponed to mid-2009 from YE08) and downward revisions in stated nameplate capacity (most notably by Natureworks).

EXHIBIT 3: BIOBASED PLASTICS (# OF PLAYERS, TOTAL CAPACITY (’000 TPY) AND TOTAL PHA CAPACITY (’000 TPY) BASED ON COMPANY ANNOUNCEMENTS

50 51 5360 61

67 68 69 69 69

0102030405060708090

2004 2005 2006 2007 2008E 2009E 2010E 2011E 2012E 2013E02004006008001,0001,2001,4001,6001,800

# of companies (LHS) Total ('000 tpy, RHS) PHA ('000 tpy, RHS)


Industrial Biotech



Page 101 of 212

Key players: While data is sketchy, industry observers estimate there are roughly 60 producers of biopolymers, not all of which are biodegradable. Earlier this decade, PLA emerged as the dominant biodegradable plastic technology due to NatureWorks’ $300m 140,000tpy PLA facility (using 0.5% of the U.S. corn crop), initially funded through a JV between Cargill and Dow Chemical. While Dow Chemical exited the JV in 2005, Cargill invested more than $750m in Natureworks, including basic R&D. In October 2007, Cargill sold 50% of Natureworks to Teijin, with an eye to further expanding capacity. Natureworks recently confirmed that effective capacity was only 70,000tpy, but insists that it should reach the initial nameplate capacity target by the end of the decade. Other key players include Metabolix (50,000 tpy PHA plant under a JV with ADM coming onstream in 2009), Novamont (35,000 tpy of a starch-blend called Mater-Bi, with another 15,000 tpy planned), BASF (14,000 tpy using a starch-PLA blend called EcoFlex, with an expansion to 60,000tpy by 2011) and Plantic (8,000 tpy of corn-based plastic).

New entrants: Biobased and biodegradable plastics are attracting a wide range of new entrants. The most notable include Dow Chemical (350,000–385,000 tpy of bio-based polyethylene by 2011), Braskem (200,000 tpy of bio-based polyethylene by 2010), Bio-On (10,000 tpy of PLA and PHA by 2009), Livan (50,000 tpy of corn-based packaging materials by 2009); Tianan Biologic (10,000 tpy of PHBV by 2009 and 60,000 tpy by 2011), and DSM (50,000 tpy PHA plant in China by 2009). Even Total has entered the PLA market with an initial target of 3,000 tpy of capacity by 2011.

EXHIBIT 4: BIOBASED PLASTICS SHARE BY REGION (%), 2006E–2013E

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2006E 2007E 2008E 2009E 2010E 2011E 2012E 2013E

Europe Asia U.S. Latin America


The new entrants are leading to a significant shift in the geographic diversity of production, with significantly more capacity coming onstream in Asia and Latin America. Indeed, we estimate that Europe’s share of bioplastic production should be roughly 23% by 2013, down from 59% in 2006, whereas Latin America should contribute 40% of total bioplastic by 2013 from zero currently.

EXHIBIT 5: BIOBASED PLASTICS CAPACITY (’000 TPY AND % OF TOTAL)

2006E % 2009E % 2013E %Natureworks 70 36% Braskem 205 25% Dow 350 23%Rodenberg 40 21% Novamont 80 10% Braskem 250 16%Novamont 20 10% Natureworks 70 9% Metabolix 200 13%National Starch

20 10% Metabolix 50 6% Natureworks 140 9%

BASF 14 7% Toyota 50 6% Novamont 80 5%Other 30 15% Other 350 43% Other 501 33%Total 194 100% 805 100% 1,521 100%Biodegradable 184 94% 472 58% 778 51%

% Metabolix 11% 26%Source: Jefferies & Company, Inc. estimates, company data

Potential demand: Industry observers have estimated the addressable market opportunity at 4.5m tpy, or 5.8x 2009E capacity, though market penetration is hindered by the need for tailored product formulations, processor partnerships and end-consumer education. For example, Europe’s COPA and COGEPA have estimated the

Industrial Biotech



Page 102 of 212

addressable market at 2m tpy: 450,000 tpy for catering products; 400,000 tpy vegetable packaging; 400,000 tpy foil packaging; 240,000 tpy for 100% biodegradable diapers; 200,000 tpy tire components; 130,000 tpy biodegradable mulch foils; 100,000 tpy organic waste bags; and 80,000 tpy biodegradable foils for diapers.

2. How are bio-based plastics made?

• Polyolefins such as HDPE, LDPE, PP, and PS, are not biodegradable because of the strong carbon-carbon bonds that form the spine of the polymer. PET is also not biodegradable, because its phenyl ring blocks degradation by water molecules (i.e., hydrolysis). Together, these plastics represent roughly 90% of all plastics. Specialty polymers such as copolyesters, polyamides, and polyethers have weaker heteroatoms (non-carbon atoms) in the spine of the polymer, making them easier to break. Poly (vinyl alcohol) can be degraded by water due to the alternation of hydroxyl groups on the carbon atoms. Bio-based plastics, in contrast, are typically made through one of the following strategies:

• Extracted and used without modification. Starch is the most common example, being competitive with polyethylene on price. The main limitation is the need for blends to reduce sensitivity to water. Synthetic blends of polyethylene and starch are only partially biodegradable (the microbes eat the starch, but leave polyethylene flakes behind). Stiffeners (limestone, cellulose from recycled paper) can also be added to increase toughness. Metabolix’s innovation is to discover how to make PHA polymers and co-polymers directly in plants, which should be price-competitive with commodity plastics while providing a far broader range of performance characteristics than starch.

• Broken down to sugars (monomers). Converted to new monomers, and then turned into polymers via chemical methods. This is the method used to make PLA, which degrades by hydrolysis. Braskem uses a chemical method to convert ethanol to polyethylene (a process it claims captures 2.5 t of CO2/t of PE, vs. emissions of 3.5 t of CO2 under petrochemical routes).

• Broken down to sugars (monomers). Converted to polymers via fermentation. This is the approach Metabolix is taking in its JV with ADM. Microorganisms accumulate the PHA, which can eventually constitute as much as 80%–90% of dry weight. The PHA is then removed with a solvent. P&G and Kaneka have a JV focused on some grades of PHAs, but commercialization appears to have been delayed due to a mismatch in their interests (consumer products vs. durable goods).

• Converted to monomers, which are then sold to other companies to convert to polymers. Some companies have started exploring the production and sale of lactide, the monomer used to make PLA, so that downstream companies can formulate their own PLA blends. Purac, for example, aims to build a 65,000–90,000 tpy lactide plant in Thailand, due to come onstream in 2009. Similarly, DuPont’s 50,000 tpy bio-PDO plant provides buildings blocks for a range of chemicals.

• Small molecules (oils) are extracted, then cross-linked using chemical methods to make thermosets. Examples would include the soy-oil based composites for John Deere tractors that are sold by Ashland (ASH, $27.61, Buy, $51 PT), as well as products by BASF and Arkema. This is, in our view, the preferred approach for producing bio-based durable thermosets. Natural oil polyols also provide renewable content for elastomers, surfactants, polyurethanes and foams. For example, some studies have suggested that as much as 40% of MDI can be replaced by natural oil polyols, or 25% of the final foam’s total weight, without impairing performance properties.

Industrial Biotech



Page 103 of 212

EXHIBIT 6: CLASSES OF BIODEGRADABLE PLASTICS

Natural/Synthetic Renewable? Manufacturer

Aromatic Polyesters Aliphatic-aromatic copolyesters Synthetic No BASF, Eastman Modified PET & derivatives Synthetic No DuPont Aliphatic Polyesters Polybutylene Succinate Synthetic No Showa Denka, SK Chemicals Polycaprolactone Synthetic No Dow, Solvay, Daicel PHA Natural Yes Metabolix, Kaneka/PG PLA Synthetic Yes Cargill, Hycail, Mitsui, Toyota Others Polyvinyl Alcohol Synthetic No Various Starch-based Polymers Natural Yes Novamont, Rodenburg, Plantic, Biop Cellulose Acetate Natural Yes Innovia Films, FKuR Ethanol-to-Polyethylene Natural No Braskem

Source: Environment Australia, European Bioplastics

3. Back to the 1980s with ethanol-based polyolefins?

Over the last three quarters, Dow, Braskem and Solvay have announced plans to build roughly 350,000 tpy of capacity for making various grades of polyethylene from ethanol (ethanol to ethane to ethylene to polyethylene). In a sense, this is a return to the 1980s, when the Brazilian government supplied subsidies for Braskem, Dow and Solvay to produce roughly 150,000 tpy of ethanol-based ethylene for making polyethylene and PVC. These prior initiatives ended, along with their subsidies, once oil prices declined in the early 1990s. With Solvay investing $135m in 60,000 tpy of ethanol-based ethylene capacity (on-stream in 2010) and Braskem investing $150m in a 205,000 tpy bio-polyethylene plant (on-stream by YE09), we estimate Dow’s JV with Crystalserv could cost $300–$500m.

4. Why Metabolic Engineering? Fermentation vs. the Foie Gras Method.

Microbial physiology, and the ensuing production of desirable value-added metabolites, can be stretched way beyond “wild type” parameters through the selective use of genetic engineering and culture conditions (referred to as metabolic engineering).

Microbial fermentation techniques have a recognized place in the production of low-volume, high-value products (like drugs), but moving to lower value, higher volume chemicals requires moves to maximize efficiency and minimize costs and waste by-products. Microbial processes are composed of a number of research, development, scale-up and commercial elements; however, they need to be viewed as an integrated whole in order to optimize yield, productivity and competitive cost comparisons. In particular, the choice between fermenting monomers and polymers is critical in evaluating the capital costs of different bio-based chemical production methods.

A fermentation vat is a complex world: aqueous mixtures of cells jostle with soluble extra-cellular products, intracellular products, converted substrate and media components. The actual growth conditions and downstream separation techniques depend highly on the location (extra-/intra-cellular), size, charge and solubility of target products; for commodity chemicals the focus is on finding the cheapest route. However, regardless of the product the highest purity in the final “broth” is the most desirable—for chemicals, fermentation represents a large percentage of the total production costs (and where “integrated” development can play a most valuable part).

Most bio-based plastics use chemical biological processes, usually fermentation, to create the monomer (picture small beads of plastic), and traditional chemical processes to make the polymer (stitch the beads together into a sheet or film). The plants provide the feedstocks, whether as starch (storing carbohydrates in corn, potatoes, wheat, etc.), protein (e.g., in soybeans), cellulose (the cell walls in wood, cotton, corn, wheat, etc.), or lipids (e.g., in vegetable oils). A clear illustration of this approach is Braskem’s “green plastic,” where sugarcane ethanol is transformed into ethane, and then converted via traditional petrochemical processes into polyethylene.

In the fermentation approach, the limiting factor is the biological organism’s ability to tolerate the alcohol it produces. Eventually, the organism kills itself. An alternative approach is what we characterize as the “foie gras” method: modifying metabolic pathways so that organisms make the desired polymer inside, like beads of fat. More technically, many microbes have the facility to produce polyhydroxyalkanoates (PHAs), which are accumulated as intracellular carbon and energy storage materials. In this approach, the technical challenge is to dismantle the feedback loops so the organism does not realize it is gaining weight. The appeal is that one can eliminate process steps and make polymers directly, rather than making monomers to be processed later on. This should significantly reduce both waste products and capital requirements: picture compressing an ethylene cracker and polyethylene plant inside the tip of a blade of grass.

Industrial Biotech



Page 104 of 212

5. Is biobased the same as biodegradable?

Using renewable resources does not guarantee the product is biodegradable. Many biorenewable plastics are not biodegradable, are only partially biodegradable, or only biodegrade in specific environments. Moreover, some of these biodegrade in part due to hydrolytic breakdown, effectively melting like the Wicked Witch of the West. This means that these polymers require further modification or additives to operate in the presence of liquids or high humidity. Other synthetic degradable plastics are actually photodegradable: they break down after several days’ exposure to sunlight into smaller, non-degradable, flakes that then form a permanent part of the ecosystem. Similarly, some polymers are only “bioerodable,” as they break into smaller flakes on exposure to heat or UV radiation, but are not necessarily digestible by microbes.

Industrial Biotech



Page 105 of 212

EXHIBIT 7: PHA RATE OF BIODEGRADATION VS. PEERS (IN MICROGRAMS/SQ. MM)

PHA PCL Cello-phane PHA PCL

Cello-phane PHA PCL

Cello-phane

Days Marine water Marine sediment Soil Soil Soil 14 27 7.5 11 11 1 11 18 2 8 28 24 27 11 17 2 11 28 12 10

42 Gone 27 11 31 3 11 48 12 11

56 - 27 - 32 3.5 - 77 12 1270 - 27 - - 8 - - - -


EXHIBIT 8: COMMERCIALLY AVAILABLE BIODEGRADABLE AND COMPOSTABLE POLYMERS*

Material Type Supplier/ Distributor ProductsDegradation

ProductsExtent of

DegradationStandard

Met

Biomax™ aliphatic copoly-esters, modified PET

Dupont/ www.allcompost.com

Coating and film for food packaging, sandwich bags, utensils, fibers.

Carbon dioxide, water, biomass.

2 to 4 months in compost depending upon temperature

ASTM D6400

Biopol™ PHB/V polybuty-rate and valeric acid

Metabolix Inc/ Biocorp Consumer disposables, Containers, trash bags, packaging

Carbon dioxide, water.

20 days in sludge, to 1 month in compost

ASTM D6400, EN13432

Eastar Bio™

Biodegrad-able copolyester

Eastman Chemical Company/ Farnell Packaging Biodegradable Products

Trash bags, film, liners



ASTM D6400, EN13432

Ecoflex™ Aliphatic-aromatic Polyester

BASF/ www.allcompost.com

Compost bags, trash bags, carrier bags, fruit and vegetable bags.



ASTM D6400, EN13432

Mater-Bi™ 60% starch and 40% polyvinyl alcohol

Novamont/ BioBag Corporation

Trash bags, lawn and garden bags



ASTM D6400, EN13432, BPI

Nature- Works™

Polylactic acid (PLA)

Cargill Teijin/ Biodegradable Food Service, Eco-Products, Inc.

Clear cups, clamshells, salad bowls

Carbon dioxide, water


ASTM D6400, EN13432

Plantic™ Starch-PVOH

Plantic Technologies of Australia/ same

Rigid containers, trays

Carbon dioxide, water.


EN 13432

Source: California IWMB

Compostable plastics can help reduce the environmental burden of plastic. For example, in 2003, plastics represented ~10% (by weight) of materials in California’s waste stream. Exhibit 8 indicates the results of published research in California. The ASTM D6400 standard differentiates between biodegradable and degradable plastics (the European standard is EN 13432). The BPI certificate demonstrates that the material meets the requirement of ASTM D6400 and will biodegrade during municipal/commercial composting.

6. Can biodegradable plastics compete with petrochemical plastics on cost, capital intensity?

While biodegradability is an attractive feature, the first hurdle is cost. Importantly, rising oil prices and advances in production processes have brought biodegradable plastics such as PHAs, PLAs and PCL within sparring distance of oil-based commodity plastics. Exhibit 9 provides a comparative overview of estimated production economics and projected cash returns at current levels. Besides presenting the capital costs for the plastic facilities, we also provide estimates for an “integrated capital cost,” factoring in the capital required to produce the key feedstocks as

Industrial Biotech



Page 106 of 212

well. To a certain extent, this is a fuzzy metric, as we do not attempt to calculate the all-in capital cost back to the refinery or oil well stages, but we believe this is roughly comparable to the capital cost of the ADM-Metabolix PHA facility, excluding the capital costs for corn cultivation, collection, and conversion into PHA feedstock.

EXHIBIT 9: COMPARATIVE ECONOMICS OF METABOLIX PHA VS. COMMODITY PLASTICS

$/lb Price Cash Cost Margin

Invested Capital

Integrated Capital

Cost Cash

Return Integrated based on ABS $1.28 $1.21 $0.07 $0.74 $1.18 6% Acrylonitrile, Butadiene, Styrene, ABS HDPE $0.97 $0.74 $0.23 $0.51 $0.91 25% Ethylene, PE Ethylene $0.40 PP $0.91 $0.65 $0.26 $0.38 $0.51 51% Propylene, PP PET $0.90 $0.86 $0.04 $0.32 $0.82 5% Ethylene Glycol, PTA, PET PS $1.11 $0.93 $0.18 $0.32 $0.94 19% Ethylene, Benzene, Styrene, PS Polycarbonate $1.85 $1.50 $0.35 $1.40 $1.97 18% PLA $0.95 $0.75 $0.15 $0.97 $0.97 15%PHA $2.50* $2.10 $0.40 $1.90 $1.90 21% Non-integrated, less robust

PHA** $2.50* $1.85 $0.65 $1.81 $1.81 36% ** Vertically integrated (MBLX) Source: Jefferies & Company, Inc. estimates * Tianan already sells PHBs for roughly $2/lb.

7. Can biodegradable plastics compete on a life-cycle basis? EXHIBIT 10: LIFE-CYCLE ENERGY REQUIREMENTS FOR PHA VS. OTHER PLASTICS

Megajoules/kg Energy Raw Materials Total Kg of oil/natural gas used to make 1 lb of plastic Thermoplastic starch 25 25

PVOH - - 57PLA 56 - 56 1.9

PHA (ferment) 11 47* 58 0.25 PET 38 39 77 2.26 PCL - - 77

HDPE 31 49 80 2.2 PS (general purpose) 87

Nylon 81 39 120 4.7 Source: Jefferies & Company, Inc. estimates, Environment Australia, Scientific American, European Science and Technical Observatory * Renewable (crop-based)

As companies increasingly look to reduce their overall environmental impact, we expect this to favor biobased plastics. For example, from a life-cycle perspective, we estimate Metabolix’s manufacturing process uses roughly half the amount of energy inputs as a commodity plastic such as HDPE. Importantly, more than 80% of this is actually renewable inputs (i.e., the crop) rather than fossil fuels. As a result, making PHAs uses almost 90% less oil or natural gas as a commodity plastic. Similarly, we estimate PHAs generate less than half of the CO2 emissions of producing a similar weight of PE or PET, or less than 1 kg/kg of plastic produced. Total GHG emissions can be reduced further by using corn stalks and windpower to power production facilities.

EXHIBIT 11: ESTIMATED COST STACK FOR PRODUCING PHA PLASTIC ($/LB)

Sweetener (3lb) $0.33 Solvents $0.10 Other Variable Costs $0.11 Fixed Costs $0.40 Total Cash Cost $0.94 Depreciation ($/lb) $0.13 Total Cost $1.07


Another way to look at this is that, fundamentally, the opportunity to displace conventional plastics hinges on both the premium customers are willing to pay for biodegradable properties and, more importantly on the arbitrage between biological feedstocks (sugar, corn) and petrochemical feedstocks (oil, natural gas) on process economics, as sources of carbon, and as producers of CO2. Moreover, biodegradable plastics will have to compete with biobased plastics and other crop derivatives (such as fuel alcohol), so process economics should take into account

Industrial Biotech



Page 107 of 212

substitution opportunities as well as the opportunity to displace petrochemicals. There are a few structural issues which frame this competitive landscape:

• Arbitraging sources of carbon. The chemical industry’s transition from wood and coal to oil and natural gas in the 20th century was driven by the relative ease of turning ethane and methane into polyolefins. To use coal or carbohydrates, in contrast, one faces two challenges: the feedstock is lower energy; and the plant will typically have higher capital costs.

EXHIBIT 12: ALTERNATIVE FEEDSTOCK OPTIONS: SWITCH TO COAL OR STARCH INCREASES ENERGY COSTS

Source: Dow Chemical

• Arbitraging variable costs. The following chart shows, as a theoretical assumption, the variable cost in terms of energy of making polyethylene from various feedstocks, including natural gas, coal, corn and biomass. In practice, experienced plant operators such as Dow Chemical have been able to extract only about a quarter of the theoretical value gap between polyethylene and the feedstocks—and this is only in traditional polyolefins where it has had decades to perfect the chemical process. Importantly, Brazilian ethanol and biomass could prove to be long-term sources of chemical feedstocks, whereas corn-based ethanol is prohibitively expensive at current levels. This provides theoretical support for announcements by Dow and Braskem of projects to produce bio-based plastics from Brazilian sugarcane. The chart below also shows where the relative costs are for electricity and gasoline, which are easier to transport (and hence more attractive) than energy for chemical plants.

Industrial Biotech



Page 108 of 212

EXHIBIT 13: ARBITRAGING THE VARIABLE COST OF ENERGY

Source: Dow Chemical, 2Q07

• Arbitraging capital costs and CO2 emissions: While coal may be attractive from a variable cost standpoint, capital costs are high (as much as $2–$3bn), whereas ethanol plants are relatively capital light (as much as 80%–90% less). A related issue is the volatility in variable costs, with significantly more volatility in ethanol and naphtha. If CO2 shifts from being an external cost to an economic cost, it will have a significant impact on the relative appeal of different feedstocks. Importantly, Dow noted that even for ethanol CO2 costs could be an issue, as fermentation typically produces 1.6lb of CO2 for each lb of ethanol produced. As a result, even if governments start charging for CO2, Middle Eastern ethane and natural gas will probably remain the most attractive feedstock for most non-biodegradable/green-branded chemicals and polymers.

Industrial Biotech



Page 109 of 212

EXHIBIT 14: RELATIVE APPEAL OF DIFFERENT FEEDSTOCKS BASED ON BOTH VARIABLE COST AND CAPITAL


EXHIBIT 15: IMPACT OF CO2 COSTS ON RELATIVE COSTS OF DIFFERENT FEEDSTOCKS


Industrial Biotech



Page 110 of 212

8. What drives market adoption?

In our view, biodegradable plastics will be adopted fastest when they are competitive on cost and performance, contribute additional environmental benefits, reduce waste, and provide additional savings of labor or energy. The most attractive opportunities include the following:

• Coated paper for products such as food wrappers and disposable cups.

• Films for packaging such as shopping bags, fresh food wrapping, plastic wrap for catering, and even compost bags and food scrap bags for municipal waste processing.

• Agricultural mulch film. PHA film could be ploughed into the soil, rather than removed, and have the potential to enhance topsoil rather than hasten erosion.

• Consumer packaging materials, such as straws, six-pack rings, bottle caps.

• As filler in wood-composite materials (in excess of a 17bn lb market, currently 30%–50% plastic filler).

In many cases, the regulatory tide has already started to move in favor of degradable plastics. For the last decade, several U.S. states have required six-pack rings be photodegradable, and more recently federal procurement programs have identified several categories where biobased products are designated as preferred, including adhesives, insulating foam and construction panels, biodegradable containers and films, carpets, glass cleaners, greases, and metalworking fluids. Unlike biofuels and other forms of alternative energy, however, bio-based and biodegradable plastics have not yet received a sustained over-arching government mandate to help create the market.

With plastics representing 5% of fossil fuel consumption (the single largest source of demand after energy and transport), we believe government intervention is inevitable. Indeed, one of the longer-term impacts of the latest effort by the European Union to evaluate the toxicity of 30,000 chemicals (REACH) could be to increase the relative cost of conventional plastics, particularly those with high additives content and known persistent pollutant emissions (e.g., PVC). France could require the use of biodegradable plastics for packaging film by 2010 and other countries could follow suit. Italy has mandated the use of biodegradable materials in two-handle shopping bags by 2010, whereas Belgium has placed a 300% tax on nondegradable bags. The Japanese government, meanwhile, has set a 2020 target that 20% of all plastics consumed in Japan would be renewable, representing roughly 3m tpy of potential demand. It Toyota alone were to attain such a target, it would represent 360,000 tpy of incremental demand for bio-based plastics.

In the meantime, as companies such as ADM and Cargill build larger facilities, and achieve economies of scale, the addressable market for bio-based and biodegradable plastics should continue to expand. Pull-through demand is also being created through initiatives by downstream corporations such as Wal-Mart (WMT, $59.73, Buy, $70 PT) to become more environmentally friendly. Wal-Mart, for example, plans to use roughly 114m PLA containers a year, displacing the equivalent of 800,000 barrels of oil. Similarly, Sainsbury has moved to cut by 50% the amount of plastic packaging used for vegetables and fruits, effectively switching 3,500 tpy of plastic demand to PLA plastic derived from non-GMO corn. Marks & Spencer aims to reduce CO2 emissions to zero by 2012, and has started to shift to bio-based packaging materials as part of this initiative through charging 5 pence for each plastic bag.

EXHIBIT 16: KEY POLYMERS PRESENT LONGER-TERM ENVIRONMENTAL RISKS

Polymer Applications Environmental issues Polycarbonate Baby bottles, sports water bottles Can leach bisphenol A, which disrupts the endocrine system

Polystyrene

Single-use disposable cups and containers, meat trays, egg cartons,

cutlery, foam for insulation and packaging

Toxic when burned. Contains styrene (affects the reproductive system), benzene, and butadiene

Polyvinyl Chloride

Building pipes and roofing, siding and flooring, window frames, also credit cards, shower curtains, beach balls,

food containers

High chlorine and additive content (e.g. phthalates), other persistent organic pollutants

Source: Healthy Building Network’s Guide To Plastic Lumber

9. What objections do recyclers raise?

Bioplastics producers such as Metabolix need to manage their brands carefully if their materials are to avoid the opposition PLA has encountered in recent quarters. Currently less than 10% of plastic waste is recycled, and margins remain thin. Recyclers have objected increasingly that PLA’s recyclability is not fully demonstrated, and they resist having PET and PLA blended in the same products. Sorting the two requires infrared sorting equipment that is too expensive for smaller recycling operations, and recycled PET is already established as a viable

Industrial Biotech



Page 111 of 212

commodity (worth $0.15-plus/lb). At the same time, PLA, having been marketed as biodegradable, breaks down quickly only under ideal (i.e., industrial) composting conditions, when microbes can digest the plastic in a 140 degree F environment for more than a week (there are only 113 in the U.S.)—and there is also some debate as to whether the lactic acid in PLA short-circuits the composting process. Also, there appears to be a growing recognition that, for the plastic that ends up in landfills (and plastics contribute 25% of landfill volume currently), biodegradability is a moot point, as that environment lacks the microbes, water, and oxygen needed to break down the plastic.

10. What are PHAs? A flexible class of natural plastics.

Much as humans store excess energy from their diets as fat, many microbes store excess carbon and energy as PHAs, or polyhydroxyalkanoates. Unlike most other bio-based or biodegradable polymers, these are literally “natural plastics”: candy for (other) microbes, they biodegrade in environments where even most “biodegradable” plastics remain inert.

EXHIBIT 17: PHA MOLECULAR STRUCTURE

O

OHH

O

R

xn

( )O

OHH

O

R

xn

( )Source: Company information * R = carbon chain up to C13, x = 1-3. Metabolix PHA is based on R=methyl, x=1. PLA is based on R=methyl, x=0, etc.

More technically, Metabolix PHA natural plastics are semicrystalline thermoplastics. The basic homopolymer is poly-(3-hydroxybutyric acid), or PHB. The underlying structure is similar to other plastics such as polylactic acid (PLA) and polycaprolactone (PCL). Metabolix can modify the plastic’s hydrophobicity, tensile strength, transition temperatures, and level of crystallinity, effectively making molecules that deliver a wide range of properties, comparable to everything from rigid thermoplastics to thermoplastic elastomers, as well as forms useful in waxes, adhesives, binders, and solvents. In some cases, the properties are even competitive while eliminating the need for traditional additives such as plasticizers, for example simply by adjusting the relative mix of hydroxyvalerate (PHV) and hydroxybutyrate (PHB) in the PHBV co-polymer.

In general, PHAs are thermally stable under 180°C and demonstrate good resistance to hot liquids. Metabolix believes that PHAs could eventually provide a viable alternative, at least on a performance basis, for more than half of the plastics market, though we believe that for many applications the world does not need another source of plastic except for feedstock diversification. The company has already conducted proof-of-concept trials to show that PHAs can be processed using existing commercial processes. In practice, however, we expect Metabolix to compete first to displace conventional plastics in select niches within the roughly 20% of the market constituted by disposable products. Currently, for contrast, the entire global bio-based plastics market is less than 1% of the total global plastics market, with demand for bio-based plastics growing faster than 15% per year.

EXHIBIT 18: PHA WATER BARRIER PROPERTIES VS. OTHER PLASTICS (IN G/M2/DAY AT 23 DEGREES C, 90% HUMIDITY, 50 MICRON FILM)

Metabolix PHAs 20-150 Annealed PLA 3,400 PCL 3,600 Ecoflex (BASF) 3,400-3,600 Bionolle (Showa Denka) 6,600 Cellulose Acetate 58,400 Polypropylene 3-5 PET 10-15 Nylon 6 15


11. How hard is it to develop an integrated PHA fermentation process?

Natural organisms can produce PHAs via fermentation, but the process has tended to be cost-prohibitive due to low yields, slow growth rates, instability, and the difficulty of isolating the PHAs. The challenge is to fine-tune metabolic pathways, with several genetic changes integrated in the chromosome in a stable fashion, expressed in a

Industrial Biotech



Page 112 of 212

coordinated fashion, and each one several times more productive than the naturally occurring variant. To develop an integrated large-scale fermentation process, for example, would take several steps:

1. Organism selection: Many microbes accumulate PHA (a range of polymers, from poly-3-hydroxybutyrate, PHB, through poly-3-hydroxyoctaoate). However, PHB is not a normal metabolite of Escherichia coli (E.coli), a selected production strain, but can be “forced” into manufacturing PHA through metabolic engineering techniques (see next). What can be improved at this stage include basic microbial characteristics, including:

• substrate versatility, • by-product formation (what, how much), • health status (i.e., susceptibility to process snafus) and • physiological make-up (maximal growth rate, aeration requirements, etc.)

2. Metabolic and Cellular Engineering: There are believed to be two principal biosynthetic pathways producing microbial PHA; including those with short chains, like PHB and others with longer side-chains (with more elastomeric properties). A combination of the two pathways probably provides the hydroxy-acid monomers—useful co-polymers of this type, defined by Metabolix as PHB-co-HX, include PHB-co-3-hydroxyhexanoate. The genes essential for this activity (understood to be currently up to nine) can be engineered into the production strain; gene “cassettes” that could be used for this purpose have been identified in a range of bacterial species (notable examples are Ralstonia eutropha and A.vinelandii). The ultimate goals of this stage are to improve or add some (or all) of the following:

• existing properties of the microbe (bias carbon flux to PHA biosynthesis), • introduce novel functions (includes ‘switching’ carbon flux to PHA, enhanced product recovery, broaden

substrate and product ranges), • fermentation under non-standard conditions (adapting process to scale-up constraints).

3. Fermentation process development: Much of the preparatory work (from academic and pre-competitive research) needs to be tuned to the commercial process, particularly in respect to microbe culture conditions and growth media optimization (this may require a significant move from a complex to a more defined medium). Looking downstream, to improve efficiency and cost factors, product recovery and purification are keys areas of focus; this can relate back into optimizing growth and media conditions, in terms of, for example, minimizing by-product formation or developing a high-cell density production method (for example, fed-batch vs. batch).

4. Downstream processing: The importance of this step — and particularly how it relates to cost in production of PHAs — must not be overlooked. The right technique at the right time; for example, Metabolix has optimized PHA recovery through incorporating a nuclease gene in the production strain (post cell lysis helps reduce broth viscosity to aid product recovery).

In terms of types of PHA, production can be tailored to specific demand. Either PHB homopolymer or co-polymers of PHB and an external organic acid feed (e.g., B-hydroxyvalerate, HV) can produce a range of thermoplastic polymers which can be processed with conventional techniques into bottles, moldings, fibers and films.

12. What genetic modifications might be needed to optimize PHA production?

PHB/PHA biosynthesis in microbes is a complex and tightly regulated process. The modifications involved are delicate. For example, a brief summary of the process (in the model organism, Azotobacter vinelandii) is as follows:

1. 3-step synthesis: Condensation of acetyl-CoA to form acetoacetyl-CoA (step 1), which is reduced by an NADPH-reductase to produce beta-hydroxybutyryl-CoA (step 2); then polymerisation by PHB synthase to form PHB (step 3). This synthesis is under allosteric control of the step-1 enzyme (beta-ketothiolase).

2. Gene regulation: A PHB biosynthetic gene cluster, phbB/phbA/phbC, codes for the enzyme activities of a beta-ketothiolase (step 1), an acetoacetyl-CoA reductase (step 2) and PHB synthase (step 3), respectively. Expression of the genes is under transcriptional control by two overlapping promoters (themselves controlled by additional genetic elements, phbR and rpoS).

3. “Global regulatory system”: another control system to regulate biosynthesis (integrated with RpoS); kinase (GacS) and regulator (GacA) enzymes offer additional biosynthesis control through interplay with rpoS.

4. PHB synthesis under different growth conditions: in exponentially growing cells, “balanced growth” conditions produce little PHB (beta-ketothiolase activity inhibited), with low transcription of phbB/A/C caused by lack of RpoS (which also affects PHB promoter activation) and low levels of PhbR (also dependent on RpoS). However, on

Industrial Biotech



Page 113 of 212


entering a ”stationary” phase, increased transcription of rpoS and phbR stimulates transcription of phbB/A/C. In addition, the tricarboxylic acid cycle activity may slow during this phase, relieving an inhibition on beta-ketothiolase.

5. Adopting a fermentation strategy to favour PHB production: Another feature of metabolic engineering to help improve process yields is selective mutation-repression to optimize metabolite flux through the PHA pathway (e.g., alginate synthesis blockade). Further pathways to be modulated include the tricarboxylic acid cycle (TCA): for example, in Azotobacter, gene knockout of pycA can reduce oxaloacetate levels, reducing flux through the TCA cycle to citrate, helping optimize flux into PHA accumulation.

More refined modulation is aimed to “force-feed” engineered microbes on chemical feedstock, aiming to harness more specialized biotransforming genes to incorporate different co-monomers that could lead to new products with new properties. Genes from a range of microbial and yeast sources are being investigated for specific product formats; these regulatory genes include:

• gdhA (glutamate dehydrogenase); • gadA/ gadB (glutamate-succinic semi-aldehyde transaminase); • 4hbD (4-hydroxybutyrate dehydrogenase), and • 4-hydroxybutyrate-CoA transferase gene.

13. Can PHAs stretch to medical applications?

PHAs are biocompatible materials that could have important medical applications. For example, an MBLX spin-off, Tepha (private), has developed a medical grade P4HB (poly-4-hydroxybutyrate); the first product TephaFLEX Absorbable suture, was cleared earlier this year for U.S. marketing. Another PHA polymer, TephELAST (absorbable elastomer), is under development for a range of medical devices in endovascular, cosmetic and regenerative medicine applications.

Industrial Biotech


Page 114 of 212

EventThis report provides an overview of the water industry, includinggrowth drivers, challenges, and likely investment themes. Whilehighly topical due to persistent scarcity, climate change and, in manyregions, aging infrastructure, the sector still requires a bottom-upapproach to identify companies with sustainable growth profiles.

Key Points• Strong secular trends. The water sector has outperformed the

S&P 500 over the past decade in part due to strong secular growthdrivers: demand for fresh water growing 2x population growth;increasing dependence on non-renewable aquifers for incrementalsupply; ongoing industry consolidation (in regional utilities, watertreatment, and key technologies); aging infrastructure (potentiallymore than $14bn/year of required investment for drinking water inthe U.S. alone); and growing recognition that water issues are alimiting factor for a widening range of industries, including newtechnology platforms such as coal-to-olefins and biofuels. Weexpect the industry to grow 4%-6%, with industrialization in theBRIC countries driving growth rates 2x-3x higher than the industry.

• Stay selective. The water sector, broadly defined, should reach$475-$500bn in sales this year, with industrials in the sectorreaching roughly $100bn. Despite the attractive macro trends, werecommend a selective stance. Political risks, fragmented endmarkets, a scarcity of "pure play" investments, and difficulty raisingprices to consumers can undermine returns or limit the scalability ofbusiness models. The industry tends to be conservative, whichslows the rate of adoption for new technologies. Thematically, webelieve the key opportunities are in infrastructure replacement,process efficiency, and turn-key solutions. In particular, weemphasize companies with a clear path to margin expansion. Wehighlight Buy-rated Nalco (industrial water treatment and "green"building designs), Northwest Pipe (new U.S. infrastructure), andChrist Water (desalination and industrial turnkey solutions) aspure-plays, while Ecolab (metering technology quantifies watersavings in restaurants and hotels) and Monsanto ("more crop perdrop") should benefit to a lesser extent.

• Balance valuations with earnings visibility. Secular growthdrivers, droughts in several regions, and limited cyclical risk havecontributed to a sharp expansion in water valuation multiples sincethe late 1990s, with utilities and water treatment companies tradingat the most elevated relative multiples. Within the sector,profitability ratios tend to drive valuation multiples, and even naivequantitative strategies add value (e.g., low-P/E stocks havedelivered an 18% CAGR over the past decade). Accordingly, webelieve valuations should be balanced with earnings consistencyand the opportunity for favorable earnings revisions.

October 10, 2008

Clean TechnologyAlternative Solutions - Water

Clean TechnologyWater Primer

Investment SummaryIndustrial companies involved in the provision, treatment, anddistribution of fresh water stand to benefit from secular demanddrivers and, in some regions, supply constraints. Returns have beensolid: the water sector outperformed the S&P 500 by 5% per year, onaverage, since 1998. Investors can do better, we believe, with aselective approach, given the range of niches and business models.



Alex Barnett, CFA, Equity Analyst00 33 1 5343 6714, [email protected]


Overview

Water is a simple chemical, two hydrogen atoms and one oxygen, but the almost $500bn water sector is complex. Investors are often drawn in by the combination of large markets and powerful secular themes: population growth; rising water consumption per capita; aquifer depletion; increasing scarcity of supplies of fresh water; and even climate change as a recent hot topic. In practice, however, the sector encompasses a wide range of business models and end market niches, and a bottom-up approach is critical, in our view. Topical themes may not work well in practice because of political constraints or a mismatch between regional wealth distributions and water needs. This piece is intended to briefly sketch a framework for approaching “investing in water,” broadly construed: both the long-term themes and reasons to stay selective.

• Powerful secular themes: Water demand tends to grow 2x the pace of population growth. The global supply of freshwater is relatively fixed, and unevenly distributed. Surface supplies of freshwater are supplemented by aquifers, which are not renewable in a reasonable time frame. On the demand side, the early stages of industrial development and urbanization typically see water demand rise sharply, both for industrial uses and consumers—at home, for leisure activities (swimming pools, golf courses), and tourism. In the United States, for example, personal water consumption increased roughly 10x per capita in the 20th century. Reflecting these themes, the water subsectors most sensitive to industrialization in the emerging markets have done the best in recent years: filtration; water treatment; and process solutions.

EXHIBIT 1: LONG-TERM RETURNS, INCLUDING DIVIDENDS, FOR WATER STOCKS

5 Year CAGR (%) 10 Year CAGR (%)

Conglomerates 6.0% 7.1% Filtration 17.7% 8.9% Flow Technologies 26.8% 8.9% Solutions & Treatment 17.2% 7.0% Utilities 15.7% 10.1% Water Rights 27.5% -13.8% Water Average (61 companies) 19.5% 7.8% S&P 500 10.2% 2.9%

Source: Capital IQ, Jefferies & Company, Inc. estimates

• Stay selective. Despite the favorable secular backdrop, we recommend investors stay decidedly selective within this highly fragmented sector, as the times when anything water related was awarded a premium valuation appear to be behind us. Beyond our overall bias to shares trading on low multiples, we believe the best opportunities will be companies with technology that has sustainable scarcity value that enables scalable business models. We believe companies that sell water efficiency may be an attractive area for investors, particularly efficiency solutions and related high-margin consumables that can be sold to industrial companies (which can pass along the costs) or agricultural applications (which represent 70% of freshwater consumption).

• A good sector for value investors. The water sector has outperformed the S&P 500 by 490bps over the past decade and 930bps since 2003. Even though overall multiples have risen as the sector has come into favor, the sector still tends to reward value-sensitive investors: buying low-P/E stocks over the past decade, for example, would have delivered an 18% CAGR, turning $1 into $5.25 without taking into account dividends.

• Infrastructure: Focus on new projects as repairs can be minimized or postponed. Consumers and their representatives tend to delay investments in infrastructure refurbishment as long as possible, particularly in regions where the population is in decline. Utilities also face a cost-benefit analysis, as a certain level of water loss controlled by temporary patches may be preferable to the expense and disruption of replacing an entire pipe system. Indeed, in the U.S. the GAO has estimated that more than 60% of utilities run pipeline rehabilitation and replacement programs below the desired levels. With this in mind, we favor companies like Northwest Pipe (NWPX, $30.72, Buy) that benefit from the installation of new infrastructure, such as pipeline projects in Nevada and Texas—dry regions with favorable demographic trends.

Alternative Solutions - Water



Page 116 of 212

• Watch out for disruptive technologies and unsustainable trends: The world is becoming increasingly dependent on mining water reserves from aquifers to keep up with demand, which could prove unsustainable as water tables sink. This is already emerging as an active consideration for the placement of chemical facilities, for example, particularly for water-intensive projects such as coal-to-olefins (in China) or ethanol (in the Midwest). As water stress becomes more visible in areas supplied by aquifers, we expect water consumption and wastewater treatment to become a more significant cost item for industrial companies. There is also a positive aspect to this, as companies that embed water savings within a broader value proposition should be able to take share. For example, Ecolab’s recent investment in technology that extracts methane from dairy farm wastewater provides an attractive opportunity to benefit from reducing both downstream pollution and dairy farm energy costs.

• Near-term implications for companies under coverage: Among the companies under coverage, four have the most direct leverage to trends in the water industry:

− Christ Water Technology (CWT AV, €2.10, Buy): Christ Water should continue to benefit from extremely strong end markets including desalination and industrial water treatment. The pitfalls witnessed in ‘07/’08 are, however, a warning that solid secular trends cannot in themselves carry a company. Execution remains key to deriving value from good markets.

− Ecolab (ECL, $38.84, Hold): Ecolab has a direct play on water demand in institutional hotel, restaurant, and hospital markets through its small water treatment services business (1% of total sales). The real leverage, in our view, is the indirect opportunity. Using new metering systems, Ecolab has started to quantify water savings to its core restaurant customers as part of its overall selling proposition. If conservation is often the cheapest source of freshwater, Ecolab’s business model should be able to capture a share of the value.

− Monsanto (MON, $79.34, Buy): Monsanto has a long-term source of value creation, in our view, by introducing traits that can reduce farmers’ water consumption and help address related issues of farm-related water contamination. In the 1990s the agricultural biotech industry focused on finding traits that reduced the need for agrichemicals, particularly insecticides. Increasingly, the most important limiting factors for yields are the availability of water and nitrogen (either in the topsoil or via fertilizer). Monsanto can address water consumption directly through optimizing hybrids and introducing drought-tolerant corn and soy varieties, and indirectly by working with equipment OEMs to design integrated solutions that tailor irrigation to each stage of a crop’s life cycle. Water contamination, meanwhile, can be addressed by reducing the need to use fertilizer (through crops tailored to use nitrogen and potassium efficiently), which would also lead to less downstream pollution. We estimate these technologies could contribute more than $0.30 to EPS in 2015 and $0.60-$0.75 to EPS by 2020. Yield enhancement, which is typically achieved without a significant increase in water requirements, could add a further $2 to EPS by 2015 and $3.25 by 2020. Together, we estimate these technologies have an NPV of $11, or 10% of Monsanto’s current share price.

− Northwest Pipe (NWPX, $30.72, Buy): Northwest Pipe is the only national large diameter water pipe company, with 40% share. Northwest Pipe is one of the few companies within the industrial coverage universe with EPS visibility. The company has good management, a solid balance sheet, and an attractive valuation. Most importantly, it is likely gaining share as large projects are required in areas such as Nevada and Texas. Notably, these projects are not driven by short-term residential changes. The company recognizes most of its growth from new projects, and a growing number of replacement projects. The company already has two-thirds of our revenue estimate for 2008 in backlog.




Page 117 of 212

1. Where is the water?

About 1,460 trillion tonnes of water cover 70% of the Earth’s surface. However, usable freshwater is only 0.5% of the total, with the rest being non-drinkable seawater (97%) or frozen freshwater (2.5%). Turning seawater into freshwater is becoming cheaper but is still a very expensive process. Frozen water is unavailable. The remaining 0.5% is very unevenly distributed, even more so when compared to demand.

EXHIBIT 2: WHERE IS THE WATER?

Water Source Volume (m cube km)

% of total

% of freshwater Comment

Oceans 1337.9280 96.5414% - Freshwater only 35 cubic km, or 2.5%. Ocean water contains on average 35,000 ppm of salt.

Icecaps & Glaciers 24.0619 1.7362% 68.700%

Glacial ice covers 10-11% of land surface area, vs. 33% in the last ice age. If it all melted, sea levels

would rise 70m (vs. 3cm if all the water in the atmosphere rained down). In the last warm spell, roughly 125,000 years ago, sea levels were 6m

higher than current levels. Ground-water 23.3992 1.6884% - Deep aquifers can store water for millennia.

- Fresh 10.5284 0.7597% 30.100% Rainfall replaces less than 0.1% each year. - Saline 12.8708 0.9287% -

Ground Ice & Permafrost 0.3001 0.0216% 0.860% Lakes 0.1763 0.0127% -

- Fresh 0.0909 0.0066% 0.260% Lakes 87% of fresh surface water, swamps 11%, rivers 2%

- Saline 0.0854 0.0062% - 20% of all freshwater is in Lake Baikal, another 20% in Lakes Michigan, Huron and Superior

Soil Moisture 0.1651 0.0119% 0.050%

Atmosphere 0.0129 0.0009% 0.040%

90% from oceans, rivers, lakes, etc., 10% from plant transpiration. A cloud weighs roughly 0.6

kg/cu. meter, or 35% less than dry air. The average water molecule stays in the air roughly 10

days. Rivers 0.0021 0.0002% 0.006% Biological water 0.0013 0.0001% 0.003%

Total 1,385.8600 100.0000% - Equivalent to 1.46 trillion tons, covering 70% of the

Earth’s surface Source: USGS

The water cycle: Water circulates around the earth in a cycle, ranging from as high as 15km above the surface to as low as 5km underground. This cycle is an inexorable process driven by solar energy, air quality, and gravity. Heat from the sun warms ocean water, and some evaporates. Ice and snow can also sublimate directly into water vapor, particularly in environments with low temperatures, low air pressure, and direct sunlight. Water is also transpired from plants, a side effect of plants opening pores to obtain the carbon dioxide needed to convert solar energy into sugar—a side effect that also serves to cool the plants. Sunlight, and consequently drier air, also leads to some water evaporating from the soil. The water vapor rises into the air until cooler temperatures make it condense into clouds. As air currents carry clouds around the planet, they combine, condense, and finally fall out of the sky as rain. Most falls directly back into the oceans, and some (roughly 10%) onto land where it can flow over the ground as rivers and streams, collect in lakes, be absorbed into the ground in a dispersed fashion, or collected in freshwater aquifers. Depending on the depth, the groundwater in these aquifers can be as much as tens of thousands of years old. In colder climates, the vapor turns to snow and accumulates on ice caps and glaciers.




Page 118 of 212

EXHIBIT 3: THE WATER CYCLE

Source: USGS

Access is the key. Most human water extractions involve water that is in the “fast cycle”. This is water that moves through the water cycle quickly, rather than spending extended periods stored in ocean currents, ice caps, or underground aquifers. In practice, however, humans can only access less than 3% of the fast cycle water. Many of the largest rivers are located in inhospitable regions, including the Amazon, Congo, Orinoco, Lena, and Yenisei. In Canada, for example, more than 60% of the available surface freshwater runs north into the Arctic.

Distribution is uneven. Humans have access to roughly 1,665,000 litres/year of supply per capita, but this metric is misleading. In practice, the water is unevenly distributed across the planet, with half of the renewable supply in just six countries (Brazil, Russia, Canada, Indonesia, China, and Colombia). Arid regions, which cover 40% of the Earth’s landmass, receive only 2% of the water supply. Asia, with two-thirds of the world’s population, receives only a third of the annual supply. Emblematic of this imbalance, China supports 20% of the world’s population with only 7% of the world’s water supply—80% of that water is in the southern part of the country. To some extent, people have offset this by storing roughly 3,500 km3 in reservoirs, including more than 45,000 dams.

EXHIBIT 4: POPULATION VS. WATER SUPPLY

% of population % of water supply Europe 12% 8% North America 8% 15%South America 6% 25% Asia 60% 36%Australia 1% 5% Africa 13% 11%

Source: UNESCO, Jefferies & Company, Inc. estimates

Aquifer withdrawals unsustainable. Given the scarcity of surface water in many regions, water supplies are commonly supplemented by drawing groundwater out of aquifers. The sustainability of the groundwater supply is limited, however, as groundwater recharges very slowly from rainfall or streams and lakes that percolate through the soil. In heavily pumped or arid regions, “groundwater mining” extracts more water than the annual recharge. In the United States, for example, water is extracted from the High Plains Ogallala aquifer (the largest in the US) 7x-8x faster than it is replaced, leading some observers to estimate a useful life of less than 50 years.

Which has multiple effects downstream. As aquifers deplete, water tables are falling around the globe, with declines of more than 100 feet in parts of Texas, Oklahoma, and Kansas, and more than 500 feet in some cities in Northern China. Falling water tables lead to dry wells or higher energy costs to extract water from deeper wells, and can lead to lower lake and river levels downstream, and in extreme cases lead to rivers and lakes running dry, with the Yellow River, the Aral Sea, and the Sea of Galilee being notable examples of this trend.




Page 119 of 212

Aquifer depletion contributes to salinization. Another key issue is that fresh groundwater can form as a layer over a deeper layer of salt water. Salt water is denser than freshwater, which is why freshwater tends to flow above saltwater in the soil in a fairly stable equilibrium. If too much water is pumped out of an aquifer that is near the saline/freshwater boundary, however, the pressure differentials will shift, the boundary will move, and the wells will become saline. Saline water can be used for some applications in industrial, mining, and power applications, and saline water extractions account for 15%–17% of U.S. water consumption. For most industries, however, the higher salt content in saline water (>10,000 ppm vs. <1,000 ppm in freshwater) can reduce manufacturing yields and increase maintenance costs.

Finally, climate change could exacerbate these dynamics. Some studies argue that even a 1- to 2-degree rise in global temperatures could sharply reduce the availability of water in countries downstream due to a reduction in the supply from the melting snow and glaciers in the mountains. In the near term, however, this may be masked by more rapid melting of the snowpack that has already accumulated, with a consequent sharper decline in water availability in the middle of the century. Increased snow melt could translate into more flooding in coastal regions, making it more difficult to muster the political will to invest in infrastructure projects predicated on longer-term threats of increasing scarcity. In the US, for example, snow melt supplies roughly three-quarters of the West’s water. Over the past 50 years, however, the spring snow pack has decreased 11%–12%. Reservoirs on the Columbia River can hold roughly 30% of the annual flow, so the system lacks the infrastructure to compensate for the combination of an earlier snow-melt and a reduced snowpack.

EXHIBIT 5: GLACIAL MEAN ANNUAL SPECIFIC NET BALANCE (MM OF WATER EQUIVALENT*)

-1400

-1200

-1000

-800

-600

-400

-200

0

200

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

Mea

nan

nual

spec

ific

netb

alan

ce(m

mw

.e.)

Source: World Glacier Monitoring Service *1 meter of water equivalent = 1.1 meters of ice, so 1400 mm w.e. = 1.54 m of ice




Page 120 of 212

2. How much do we use?

Global freshwater consumption is estimated at roughly 3,760 bn m3 annually, with about 15% of that in the United States and 8% in Europe. On a per capita basis, the US uses the most fresh water, at approximately 1,730 m3/person, versus an EU average of 600 and a global average of 664.

EXHIBIT 6: WATER ABSTRACTIONS FOR PUBLIC WATER SUPPLIES, IRRIGATION, INDUSTRIAL, AND ELECTRIC POWER

Volume Per capita/ as % of internal

(m m3) (m3/capita) resources surface water (%)

Groundwater (%)

North America 591,600 1,480 10.5 76 24 OECD – All Europe 287,900 560 14.2 79 21EU - 15 222,700 600 19.4 79 21 OECD 1,017,700 920 11.5 78 22World 3,760,000 664 9.0 .. .. USA 47,799 1,730 19.4 75.9 24.1 Canada 42,214 1,420 1.5 95.6 4.4 Mexico 72,564 730 17.3 62.2 37.8 Japan 86,104 680 20.3 87.2 12.8 France 30,931 530 17.1 80.6 19.4 Germany 38,006 460 32.5 83.7 16.3 UK 12,375 230 22.2 80.8 19.2

Source: OECD

In the US and developed world in general (although there is a great deal of variation) the primary use of water is in electricity generation, where water is used for cooling. Agriculture is typically the second largest consumer of water resources, followed by domestic consumption. The inverse is seen (again, with significant variation) in developing nations where agriculture is the predominant consumer of water and industrial use remains limited.




Page 121 of 212

EXHIBIT 7: US WATER CONSUMPTION, 2000 AND GLOBAL WATER USAGE TRENDS BY INCOME LEVEL

Public Supply11.0%

Industrial5.0%

Thermoelectric Power48.0%

Irrigation34.0% Domestic

Mining, Aquaculture,

Livestock2.0%

26 %

61%67%

21 %

21%20%

12% 18%

53%

0%

20%

40%

60%

80%

100%

Lo w and middleinco me co untries

Wo rld High inco meco untries*

*with GDP per capita > 20,000 USD in 2002

Industrial

Do mestic

A gricultural

Source: USGS, Aquastat

And How Fast Is Consumption Growing?

Over the course of the 20th century water withdrawals increased over six times — almost double the rate of global population growth. While on an absolute basis Asia has been responsible for ~60% of the increase in water withdrawal, growth rates in Europe and North America have significantly exceeded those of the developing world.




Page 122 of 212

EXHIBIT 8: GLOBAL TRENDS IN WATER USAGE

0

2,000

4,000

6,000

8,000

10,000

0

1,000

2,000

3,000

4,000

5,000UrbanRuralWater Co nsumption

6.5x

4x

0

500

1000

1500

2000

2500

3000

1900 1920 1940 1960 1980 2000 2020E

A sia A fricaN. A merica S. A mericaEuro pe

Source: Jefferies Intl., Unesco, Aquastat, IFPRI

Key drivers of increased water demand have included:• World population growth. The global population increased from 1.6b in 1900 to 6b in 2000, a compound growth

rate of 1.3%. While growth is expected to slow gradually, Unesco forecasts that the global population will grow by70 million per year to reach 9.4 billion by 2050. Almost all that growth is expected to be in low-income countries.

• Economic development. While per capita water withdrawals typically stabilize and can even eventually decline asmore effort and investment is spent on using water more efficiently, the early stages of industrial developmenttypically see usage of water rise sharply for domestic and industrial purposes. Bearing out that relationship, afterrising significantly over the early part of the 20th century, overall water consumption levels in the OECD have beenroughly flat since the 1980s due to initiatives to use more efficient irrigation techniques, better process controls,and the decline of water-intensive industries such as mining and steel.

• Rising affluence. As per capita income increases, demand for water for domestic usage also increases due toshifts in lifestyle. Improved hygiene, water consumption for leisure (swimming pools, golf courses, etc.), and waterdemand for tourism are key contributors to this increased water withdrawal. As an example, US homeconsumption has risen roughly 10x over the past century, compounding the impact of population growth.

More worrying than simple growth in consumption is the underlying trend towards greater reliance on aquifer orgroundwater supplies, a non-renewable resource. Groundwater from aquifers typically contributes 21%–24% oftotal freshwater supplies, with the US at the higher end of the range, while a water-rich country, such as Canada,might use almost no groundwater (less than 5% of total consumption—although some industry observers argueaquifers play a greater role). As seen in the chart below, while overall withdrawal levels in the OECD have remainedrelatively stable since 1980, withdrawals from aquifers have continued to grow and are offsetting decreasing use ofsurface water resources.


Popu

latio

n (m

illio

n)

Wat

er C

onsu

mpt

ion

Km

3

Ann

ual W

ithdr

awal

s (k

m /

yr)

3



Page 123 of 212

EXHIBIT 9: TRENDS OECD WATER CONSUMPTION BY SOURCE

0%

20%

40%

60%

80%

100%

Surface Groundwater

90

95

100

105

110

1980 1985 1990 1995 2000 M o stRecent

Gro undwater Surface Water To tal

Source: OECD

In the developing world that trend is even more pronounced. While official statistics may be imprecise, industry observers estimate that India, China, and Pakistan, in particular, have ramped up their consumption of aquifer water, to the point that these three countries represent as much as half of the world’s consumption of aquifer water for agriculture. Saudi Arabia’s programs to grow wheat, alfalfa, and most recently cattle in the Sahara have used an estimated 60% of the water from the Arabian aquifers over the past 20 years. In our view, these regions are fated to eventually follow the experience of the US aquifers, where overpumping is leading to the porous rocks that comprised the aquifer being crushed by the earth above them, crippling the ability of the aquifers to recover.

Discussions of local water consumption, however, do not fully capture a population’s water footprint. Much as one can look at trade flows as a way to shift energy requirements globally, one can also look at trade flows as a way to shift a water footprint to another region. Studies of the water used to produce different foods, for example, suggest that a pound of beef requires as much as 5x a similar amount of eggs and 10x a similar amount of wheat. From this perspective, major exporters of water (in the form of downstream products) include the United States (a third of all its water use is exported, largely as grain or meat), Canada (grain and meat), Australia (cotton and sugar), Argentina (beef), Thailand (rice), and Pakistan (cotton, using as much as a third of the annual flow of the Indus River). While methodologies for quantifying this effect are debatable, shifts in consumption towards higher water-content goods (e.g., beef, cotton) could have a noticeable impact on demand trends in export-oriented countries.

Either way, industry projections suggest that as much as 40% of the Earth’s population could face either water stress (24%) or scarcity (18%) by 2050.


% o

f Fre

shw

ater

Abs

trac

tion

Can

ada

U.S

.

Switz

erla

nd

Fran

ce

U.K

.

Den

mar

k

OEC

D

(198

0=10

0)R

elat

ive

With

draw

al L

evel



Page 124 of 212

EXHIBIT 10: ESTIMATED POPULATION FACING SIGNIFICANT WATER STRESS (BN PEOPLE) ASSUMING BASE UNESCO ESTIMATES FOR POPULATION GROWTH*

Medium Population Projection10

9

8

7

6

5

4

3

2

1

0

1995 2000 2005 2010 2020 2025 2035 2040 2045 2050

Peop

le(B

illio

ns)

Medium Population Projection10

9

8

7

6

5

4

3

2

1

0

1995 2000 2005 2010 2020 2025 2035 2040 2045 2050

Peop

le(B

illio

ns)

Source: ITT *Gray= limited water stress, Light blue = medium water stress, Dark blue = severe water stress




Page 125 of 212

3. What’s in the water?

Water quality and safety represents, in our opinion, a sustainable growth driver for the sector. Water quality standards are driven by concerns over human safety, agricultural implications, and broader environmental concerns.

The most important driver, in our view, has been sanitation. • In the developed world, changes in regulation and increasingly strict water standards are the key driver of

demand for water treatment and can lead to step-changes in filtration technology. Of particular note is the trend towards increased regulatory targeting of contaminants that can affect human endocrine systems.

• In the emerging markets, economic development, and with it demand for cleaner water and better access to that water, will be the key driver of growth. In these markets lead, fluoride, arsenic, hydrogen sulfide, and other contaminants affect significant populations. In India and Bangladesh, for example, the population faces arsenic poisoning from water washed down from the Himalayas over millions of years and stored in floodplains. Initiatives in recent decades to drill wells to limit the spread of diseases carried in surface water have led to greater awareness of the risk from arsenic contamination in the underground water, which some studies estimate could affect the 800m people living downstream from the Himalayas.

In the developed world, we expect investors to focus on the opportunities created by existing and proposed regulations that impact water distribution systems. In the US the EPA has estimated that meeting these regulations could require $45bn in infrastructure investment over the next 20 years. Of particular concern are microbial contaminants such as giardia and e.coli. Chemical contaminants such as nitrates, arsenic, perchlorates, lead, and copper are expected to represent a growing need as well, particularly given recent EPA regulations tightening arsenic limits for water.

Perhaps the highest-profile new regulation was the introduction in 2006 of new standards for arsenic content. As of January 1, 2006, the EPA lowered the threshold to 10 ppb from 50 ppb, a standard which is more onerous in the Western states. Implementing this regulation is expected to cost $20bn, and could lead to some of the smaller municipal systems to acquiesce to privatization or consolidation if they cannot invest the necessary capital. Indeed, smaller systems are already three times more likely to violate the Safe Drinking Water Act, compared to larger systems, in part because local hurdles to implement rate increases have curtailed capital investment and maintenance spending.

EXHIBIT 11: 20-YEAR WATER INFRASTRUCTURE NEEDS ($BN) FOR CONTAMINANTS

Source: EPA

In the EU, legislation is more piecemeal than that of the United States, with member states deciding for themselves how EU targets should be met. The result is significant divergence across the member states and a time-lag when it comes to European-wide reporting.

While regulation is something of a hodgepodge across the EU, the overarching water quality regulations are the Urban Waste Water Directive, the Drinking Water Directive, and the Water Framework Directive. • The Urban Waste Water Directive was introduced in 1991 and covers urban and industrial wastewater treatment.

The directive mandated wastewater treatment across the E.U., with an end goal of near universal treatment. Estimated spending on the directive through 2010 is estimated to be >€150b, but the vast majority of the directive has now been adopted, leaving little incremental investment on the table.

• The Drinking Water Directive of 1998 laid down minimum standards for microbiological and chemical content in drinking water.

Regulations $bn Surface Water Treatment $27.53 Coliform (VOCs, SOCs, IOCs, etc.) $2.63 Nitrates and Nitrites $0.50 Arsenic $0.96 Lead and Copper $2Trihalomethanes $0.20 Others $1.33 New and Proposed Rules $9.93 Total $45.08




Page 126 of 212

• The Water Framework Directive attempts to document and set in place a management plan for all river basins in the EU. The directive comprises of an all-out prohibition on direct discharges to groundwater, and (to cover indirect discharges), a requirement to monitor groundwater bodies so as to detect changes in chemical composition, and to reverse any antropogenically induced upward pollution trend.

Because of the patchy implementation of these directives and because water supply and sanitation do not fall under the direct mandate of the EU, changes in European water quality are likely to be a more gradual process than in the U.S., where new regulations can spur significant short-term investment demand.

EXHIBIT 12: SALT SALINITY LEVELS (DS/M) AT DIFFERENT LEVELS OF YIELD LOSS (%)

Salt level at Yield Loss Crop Salt tolerance threshold (dS/m) 10% 25% 50% Corn 1.7 2.5 3.8 5.9 Cotton 7.7 9.6 13 17Rice 3 3.8 5.1 7.2 Soybean 5 5.5 6.2 7.5 Wheat 6 7.4 9.5 13 Potato 1.7 2.5 3.8 5.9 Tomato 2.5 3.5 5 7.6 Strawberry 1 1.3 1.8 2.5

Source: World Bank, 1993

Another key emerging concern regarding water quality is salt content, particularly for agriculture. Higher salt content in the soil can reduce crop yields. Indeed, industry estimates suggest that excessive salt levels affect the agricultural viability of approximately 150–200m acres (a quarter of that severely) globally. One way to gauge this is that the World Bank has estimated that salinity affects as much as 28% of the irrigated land in the United States, 23% in China, 21% in Pakistan, 11% in India, and 10% in Mexico. Other parts of the world are even more vulnerable: in Turkmenistan, for example, almost half of the irrigated land is impaired by salinity. While estimates vary, salinity appears to affect an additional 10–25m acres of arable land each year, largely offsetting the conversion of new marginal land to agriculture.

Other near-term opportunities for companies include, in the United States, MTBE remediation (potentially a $10bn-plus market opportunity) and perchlorate remediation (>$1bn). Regional water quality standards in the emerging markets should also drive a significant opportunity over the next ten years.

One long-term implication of the trend towards new water standards is we could eventually see more “tiering” of the water supply. We already see this in some industrial facilities, such as semiconductor fabs or pharmaceutical plants, where the threshold for acceptable contamination is much lower than for human consumption. Over time, utilities may be able to allocate water, depending on its quality, to different applications, optimizing water treatment costs.




Page 127 of 212

4. How much does water cost?

Recent surveys have highlighted steady increases in municipal water rates in many countries. For example, the OECD estimates that over the past five years, average municipal water rates have risen at a 4.9% CAGR in the US, 9.6% in Canada, 5.7% in the UK, 7.7% in Australia, and 8.4% in South Africa. Those rates are driven by actual cost of delivery, but are also heavily influenced by progressive and regressive pricing systems and by cross-subsidies between industrial and domestic users.

Cost of supply is determined by proximity of raw water, the degree of purification required, and settlement density. The cost of delivering water can vary based on both the cost of the water ($0.03–$0.10/m3 in the US West, as high as $0.75/m3 in Australia during the 2006 drought) and transportation costs (in some cities north of $1/m3,particularly in developing countries).

EXHIBIT 13: CONSUMER WATER COSTS IN SELECT OECD COUNTRIES (IN US$/M3)

0.0

0.5

1.0

1.5

2.0

2.5

Source: OECD

Consumer costs vary widely, mostly due to government subsidies: $0.66/m3 in the US on average, as much as $2.25/m3 in Germany, and more than $3/m3 for high-volume users in Japan. Similarly, farmers are typically subsidized. In California, for example, farmers use roughly 20% of the state’s water and pay only $0.01/m3, vs. the $0.50–$0.60/m3 that consumers pay in Los Angeles. The take-home message is that users do not bear the full costs (economic and environmental) of water supply.

EXHIBIT 14: WATER CHARGES AS % OF HOUSEHOLD INCOME AND PER CAPITA GDP, 1990S

% of household income % of per capita GDP % of full cost recovery U.S. 0.5% 1% 50% Portugal 0.5% nmf 18%Denmark 0.8% 1.60% 89% England 1.2% 2.28% 92%France 1.1% 2.20% 73% Germany 1.0% nmf 83%Greece 0.4% nmf 19% Ireland 0.3% nmf 16%Korea 0.6% 0.64% 67% Spain 0.4% 1.52% 25%

Source: CBO, OECD

5. What are the secular investment themes?

We define the water industry broadly, including water utilities, desalination companies, water treatment, and companies that manufacture and install pipes, pumps, filters, membranes, and turn-key systems. Given the paucity of pure plays in the sector, much of the investor focus has been on the most direct plays, the water utilities. Investor


US$

Pric

e of

Wat

er (m

)3

U.S

.

Can

ada

Sout

h Af

rica

Aust

ralia

Italy

Fran

ce

U.K

.

Ger

man

y



Page 128 of 212

interest, particularly since 2000, has been spurred by favorable regulatory climates, capital discipline, and ongoing consolidation prompted by the prospect of the need to implement significant infrastructure investments over the next two decades (estimated at 20x the average household’s current annual water and wastewater budget). Compelling opportunities, in our view, can also be found in derivative themes, particularly the following:

• Infrastructure components: With the United States facing significant infrastructure investment requirements over the coming decades, we favor the infrastructure companies and those that sell components as long-term plays in the sector. Northwest Pipe provides a direct play on increasing demand for new pipe to replace aging infrastructure.

• Industrial water treatment: With utilities facing political constraints on how fast they can raise prices for consumers, we expect the cost of fresh water, and wastewater disposal, to rise more quickly for manufacturers. This provides a secular growth driver for the industrial water treatment companies, particularly Nalco (NLC, $12.97, Buy).

• Agriculture: Micro-irrigation, particularly in the United States, and more water-efficient crops, globally. We would emphasize Monsanto as a long-term play on addressing agricultural water consumption (70% of all freshwater demand).

• Desalination and Purification remain sexy: Improved cost effectiveness and booming demand for water in a number of arid regions are driving growth in desalination demand to rates well above the overall water sector. At the same time, emerging purification systems which can lessen energy costs and increase throughput have an extremely receptive audience in an energy constrained world. Growth rates in the disinfection/purification subsector are expected to be amongst the highest in the industry.

EXHIBIT 15: GLOBAL WATER INDUSTRY (SALES IN $BN AND GROWTH RATES)

Business Segment 2008E Growth Comment Components & Chemicals

Water Treatment Chemicals $14 2%-5% Industrial 55% (1/2 is utilities, refineries & downstream

chemicals), municipal 35%-40%

Filtration $21 5%-8% $6bn for industrials, $6bn for municipalities. Membranes 40%-

45%. Excludes $29bn for non-water filtration. Pipes $50 4%-7% U.S. market roughly $6bn, large diameter pipe $850m Pumps $30 5%-6% 20%-25% industrial. Valves $45 6%-9% 20%-25% industrial Sub-total $160 4.7%-7.3%

Infrastructure & OperationsMunicipalities $63 2%-4% Wastewater treatment. China targets $10bn/year 2006-2010.

Industrial $59 3%-6% Growing 10%-20% in developing countries.

Excludes pumps, valves & chemicals.

Distribution $53 5%-8% 10%-15% in developing countries

Disinfection/Purification $15 10%-15% Centralized systems, ex. consumables. 80% chlorination.

50% of new sales involve membrane, ozone or UV technologies. Desalination $10 9%-14% Sub-total $200 4.0%-6.3% Other

Bottled water $95 6%-12% Risk of consumer/regulatory backlash Residential $10 5%-8% Irrigation $10 6%-8%

Other $20 3%-5% Services, consulting & testing Sub-total $138 5.4%-10.4%

Total $495 4.7%-7.9% Total Industrial $99 3.6%-6.5% 20% of total Total Municipal $227 4.5%-7.1% 46% of total

Source: “The World’s Water, 2006-2007” edited by Peter Gleick, company information, Jefferies & Company, Inc. estimates




Page 129 of 212

Utilities: Political risks, environmental challenges, aging infrastructure set the stage

The municipal water treatment market is estimated at roughly $260bn, including $6bn for filtration and purification products, versus roughly $10bn for residential water treatment and filtration applications. Key factors influencing municipal demand for wastewater and drinking water treatment equipment include population growth, consumption of potable water per capita, increasing focus on wastewater capture and treatment, and the need to either retrofit aging infrastructure or install new filtration equipment to handle the inflow and infiltration of rainwater into the municipality’s wastewater network. The combination of local monopolies, stable and growing end market demand, and favorable treatment by regulators supporting stable returns on investment (at least in the US) has led to the utilities delivering fairly stable returns, and outperformance since 2000.

• Scarcity of traded assets: The publicly traded US water utilities have a combined market cap of only $6bn, with more than half of that due to Aqua America (WTR, $14.91, NC, $4bn market cap.). The limited number of publicly traded water companies has contributed to a scarcity premium for the sector.

• Consolidation: The US water utility industry remains highly fragmented, with more than 52,000 municipal water systems and more than 15,000 wastewater systems. Roughly 81% of the population is served by only about 8% of the municipal water systems, and these 3,900 systems represent the most attractive bolt-on acquisitions due to economies of scale. In contrast, 84% of the water systems serve small communities (less than 3,300 people). Rising capital requirements due to EPA regulations, coupled with a local reluctance to raise consumer water rates, can lead to these smaller systems being amenable to consolidation by private sector utilities that can spread the cost over a larger population base. In practice, this defers, but does not eliminate, the longer-term political risk.

EXHIBIT 16: MUNICIPAL WATER SYSTEMS BY POPULATION SERVED (% OF SYSTEMS AND % OF POPULATION SERVED)

% of Systems

<50056%501-3,300

27%

>100,0001%

10,001-100,000

7%3,301-10,0009%

% of Population<5002%

>100,00044%

501-3,3007%

3,301-10,00010%

10,001-100,000

37%

Source: EPA

• Regulatory structure drives returns: Regulators set returns for the water utilities based on their rate base, orinvested capital. Investing to maintain and replace infrastructure or to expand to serve new customers provides themost direct lever for improving net income, assuming capital is allocated in a disciplined fashion. A related risk isthat the utilities need regulatory approval for rate increases even to cover the inflation in operating costs. Whileregulators have been fair, in general, in the United States, the experience in other countries provides a salutarywarning that this could change. In particular, longer-term utility investors need assurance that the utilities can passthrough to customers any incremental costs associated with shifts in the water supply mix.

• Demographics drive investment: Water infrastructure is typically sized to meet peak demand. Demand, in turn,tends to grow at roughly 3%, mostly due to population growth. The combination of significant capital intensity andlong asset lives (more than 50 years) leads to a mismatch in incentives. Multi-state utilities have an incentive tofocus on expanding in regions where population growth is expanding, as the political climate is more favorable. Instates and municipalities where the population has been shrinking, however, the required investment to maintainand replace worn-out assets can be viewed as politically unpalatable. While the EPA and CBO estimates suggestthat the average US water utility bill needs to double (as a % of income) by 2025, we expect this to be delayed.Simply put, water infrastructure repair is less rewarding, politically, than highly visible investments in highwayprojects, airports, or stadiums. After all, water investments are largely invisible to consumers unless theinfrastructure fails.




Page 130 of 212

• Growth tied to consumption. In the long run, the water utilities need to find ways to decouple their growth from the consumption of water. Currently, most water utility contracts are based on the amount of water consumed. As a result, utilities earn more during hot periods, and less when more rain falls than usual. Conservation initiatives, in this regulatory structure, represent a double threat. First, by reducing the amount of water each person uses, conservation directly attacks utility revenues. Second, conservation initiatives can lead to less water being used than was expected when water infrastructure projects were first planned, undermining the ROI on these projects.

• Managing the investment cycle. Higher spending on replacing and repairing infrastructure appears inevitable (see discussion below for full details). These infrastructure projects, due to their long horizons (as much as a century), are typically funded out of rate revenue. If the regulators fail to approve rate increases sufficient to cover the costs of these projects, however, the utilities could see more of their cash flow allocated to these projects. At the same time, every 15–20 years the utilities face a spike in capital spending for wastewater infrastructure, pumps, and other components, which would be funded externally, depending on the regulatory climate.

• Foreign firms were disappointed after large investments in the 1990s. An ostensibly favorable regulatory environment and a highly fragmented market of traditional utilities attracted significant investments in the 1990s by large foreign water companies such as Suez, Vivendi, and RWE. Returns on capital were undermined, however, by less favorable contract terms than initially expected, rising expectations for M&A multiples, and in some cases local political opposition to contracting water supplies to foreigners (particularly post 9/11). With this in mind, we believe that foreign firms are less likely to drive water utility M&A over the next decade.

• Non-regulated opportunities can alter risk profile. To improve their growth prospects, some water utilities have explored non-regulated avenues to provide services, typically through arranging for municipalities to outsource the management of their water infrastructure, rather than sell the assets outright. This keeps the liability for the longer-term capital requirements with the municipality, and theoretically allows for higher returns on capital. Stiff competition, however, can lead to subpar returns, as the lack of control on the underlying infrastructure undermines negotiating leverage with the regulators.

• History of outperformance: Since 2000, the US water utilities have outperformed the S&P 500 by 109%, or a 10% CAGR. Key drivers for the sector historically have been capital allocation, project execution, relationships with regulators (who typically authorize ROEs around 10%–14%), acquisition integration, and the appropriate handling of non-regulated “operating” contracts.

EXHIBIT 17: LONG-TERM RETURNS FOR WATER UTILITY STOCKS, ABSOLUTE AND VS. S&P 500

Source: Capital IQ

Infrastructure: Large opportunity, with multiple cross-currents

$23bn/year opportunity? Infrastructure investment has provided a sustainable opportunity for water investors, particularly the repair and replacement of aging infrastructure in the developed world. In the U.S., for example, more than a quarter of water pipes are classified as “poor” or “very poor” by the EPA, up from 10% in 1980. At current rates of replacement the EPA expects as much as 45% of the existing infrastructure to be due for replacement by 2020. Overall, a 2002 EPA study estimated that refurbishing the water infrastructure could require north of $276bn in investment by 2020 above current projected levels. Investments in wastewater infrastructure could more than double this amount. These investments would augment approximately $36bn in annual spending on drinking water and $25bn for wastewater currently. Opportunities in both areas could include suppliers of piping, valves, and filtration materials.

0

100

200

300

400

500

600

700

Jan-

85

Jan-

87

Jan-

89

Jan-

91

Jan-

93

Jan-

95

Jan-

97

Jan-

99

Jan-

01

Jan-

03

Jan-

05

Jan-

07

0

20

40

60

80

100

120

140

160

Utilities vs. S&P 500 (RHS)




Page 131 of 212

EXHIBIT 18: U.S. ESTIMATED NEW INFRASTRUCTURE SPENDING, 20-YEAR OUTLOOK ($BN 2003 DOLLARS)

Study $bn EPA, 2005 (drinking water only) $276.8

Community Systems $260.1 Arsenic $0.9 New Regulations $9.9 Other $5.8 Distribution $183.6 Storage $24.8 Treatment $53.2 Source $12.8 Other $2.3

Water contaminants $55 EPA, 1995 (drinking water) $166 EPA, 1999 (drinking water) $167 EPA, 2003 (drinking water) $277 WIN, 1999 (drinking water) $420 AWWA, 2001 (drinking water) $170 Stratus Consulting, 1998 $348.3 EPA, 2000 (clean water and watershed survey) $181.2 CBO, 2002 $245 to $424 EPA, 2002 $170 to $493*

Source: EPA, CBO, WIN * $122bn capital gap for clean water, $102bn for drinking water, with operating & maintenance gaps of $148bn and $161bn respectively.

Similar opportunities present themselves in globally. Indeed, industry estimates for global water and wastewater infrastructure needs run as high as $3 trillion over the next 20 years. Some countries are moving more aggressively on this: China, for example, targets $17-$20bn a year in its current five-year plan for upgrading its wastewater treatment infrastructure. In the OECD, expenditures on water infrastructure and operating costs are expected to rise from $576bn currently to $1,038bn in 2025. The forecast 3.5% compound growth rate is expected to be widely spread; although with faster growth in the developing nations, notably the BRICs, and in the US.

EXHIBIT 19: PROJECTED EXPENDITURES ON WATER AND WASTEWATER SERVICES – AN INSIGHT ($BN DOLLARS)

0

50

100

150

200

250

300Current

20152025

Source: ‘The Impacts of Change on the Long-Term Future Demand for Water Sector Infrastructure’ by Ashley, R. and Cashman, A., OECD.

Mismatch between demographics, political incentives, and replacement cycle. Pent up demand for water infrastructure investment is driven by a historical mismatch between demographics, the replacement cycle, the water industry’s incentive structure, and technology and environmental issues.


$bn

U.S

.

E.U

.

Chi

na

Oth

erBR

ICs



Page 132 of 212

• Demographics and the Replacement Cycle. Water infrastructure needs to be designed for peak capacity. As a result, as communities expand, the infrastructure needs to be installed ahead of the population growth. The main components of the infrastructure, however, can last as long as a century. Pipes, in particular, depending on the materials used, can last anywhere from 75–125 years. Urban communities tend to wax and wane, however, and sometimes by the time the infrastructure comes due for replacement the population has actually declined. As a result, the inflation-adjusted costs for replacement can end up being higher, on a per capita basis, than initially expected.

EXHIBIT 20: AVERAGE USEFUL LIFE OF KEY INFRASTRUCTURE COMPONENTS

Clean Water Useful Life (Years) Collections 80-100 Treatment Plants - Concrete Structures 50Treatment Plants - Mechanical & Electrical 15-25 Force Mains 25Pumping Stations - Concrete Structures 50Pumping Stations - Mechanical & Electrical 15Interceptors 90-100

Drinking Water Reservoirs & Dams 50-80 Treatment Plants - Concrete Structures 60-70 Treatment Plants - Mechanical & Electrical 15-25 Trunk Mains 65-95 Pumping Stations - Concrete Structures 60-70 Pumping Stations - Mechanical & Electrical 25 Distribution 65-95

Source: EPA

Surveys have found that, on average, roughly 60% of drinking water utilities and 77% of wastewater utilities replaced less than 1% of their pipelines each year in 1998–2000. This is a recipe for a surge in capital spending at some point in the next 20–30 years—or a decline in the safety of the water supply. Currently, the United States faces replacement costs related to pipes installed in the 1880s–1940s. Spending requirements should increase once the water infrastructure assets installed in the 1950s–1960s come due for replacement. In those two decades, water infrastructure investment ran $300–$500m/year (adjusted for inflation to 2000 dollar), a level that was only matched in four other years since 1970. Some estimates place the total replacement value of water mains at $6,300/household on average, with another $3,700/household for water treatment plants, pumps, and related infrastructure. The AWWA estimates that replacement expenditures in some cities could rise 3x–4x over the next 25 years to accommodate the aging infrastructure dating from the middle of the last decade.

EXHIBIT 21: DESIRED AND ACTUAL REHABILITATION AND REPLACEMENT RATES FOR PIPELINES (1998–2000 AVERAGE)

Italics, bold = replace less than desired Rate at which rehabilitation/replacement actually occurred (%) Desired Rate (%) 0 to 1 >1 to 2 >2 to 3 >3 to 4 >4 Total Drinking water utilities 0 to 1 percent 87 8 2 1 2 100>1 to 2 percent 63 23 5 1 6 100 >2 to 3 percent 43 33 17 1 8 100>3 to 4 percent 32 45 4 14 6 100 >4 percent 35 14 5 5 41 100Wastewater utilities 0 to 1 percent 85 8 1 1 4 100>1 to 2 percent 47 40 7 4 3 100 >2 to 3 percent 51 23 18 2 6 100>3 to 4 percent 35 39 10 4 12 100 >4 percent 28 23 7 6 36 100

Source: GAO




Page 133 of 212

EXHIBIT 22: WATER UTILITY INVESTMENT (IN $M 2000 DOLLARS) FOR 20 UTILITIES, 1870–2000

Source: AWWA

• Industry incentive structure. Federal funding for water infrastructure has declined 75% since 1980 in real terms, and now represents less than 5% of funding for water and wastewater infrastructure. This makes the utilities more sensitive to local and state-level funding cycles, which in some regions could lead to project delays due to regional balance sheet limitations. Another constraint is economies of scale: infrastructure requirements per household can be 3x–4x higher for smaller water systems, due to both a smaller operating base and, importantly, a tendency for smaller systems to have to conduct more of their maintenance spending at the same time (as the initial infrastructure wears out).

Leaks: Arbitraging water losses against replacement costs. Water infrastructure typically needs a “full fix” to significantly extend the useful life. Shorter-term solutions are available, however, through local repairs or sleeves. Average leakage rates for Public Water Supply (PWS) range from 10% in Austria, Australia, and Denmark to 33% in the Czech Republic (OECD 1999). In the United Kingdom in the 1980s some 30% of all water was lost from water distribution. In some emerging economies as much as 50%-60% of treated water is lost to leaks. Even in these environments, however, reluctance on the part of regulators to allow increases in water prices can slow the utilities’ refurbishment cycle.

• Environmental/Political/Technological issues. During the 20th century, the three main drivers behind the expansion of water infrastructure were population growth and increased per capita demand, the increase in irrigated culture, and industrial development. For the 21st century, what will increasingly spur spending on water infrastructure will be a combination of heightened environmental and political awareness (driving regulation of wasteful practices in industry). Technological advances will be crucial in responding to strains on infrastructure, and at the same time balancing environmental (or ecological) considerations for the viability of local, regional, and global water systems.

An investment cycle in the emerging markets highlights a global opportunity: Even as the developed world faces the need for an infrastructure maintenance cycle, the emerging markets are building out their water infrastructure to keep pace with industrialization. Urban projects are complemented by infrastructure to transport water to where the cities are. In northern China, for example, canals under construction will eventually convey water from the Yangtze river 1,200 km to Beijing and other cities.

Water-metering could provide better leverage. Smart meters, coupled with the appropriate pricing strategies, can reduce consumer water use by as much as 20%. We believe this presents two opportunities. First, companies that can deliver turn-key solutions for water metering should benefit from an ongoing upgrade in the distribution infrastructure. Second, software companies may be able to benefit by delivering tailored solutions that help the utilities charge consumers according to both the volume of water used and the application. At this point, we would favor companies that can deliver “prices to devices,” such as technologies that might help a building optimize the water used for dishwashers or sprinkler systems based on the costs charged by the utilities. Consumers, in our




Page 134 of 212

1998

1994

1990

1986

1982

1978

1974

1970

1966

1962

1958

1954

1950

1946

1942

1938

1934

1930

1926

1922

1918

1914

1910

1906

1902

1898

1894

1890

1886

1882

1878

1874

1870

0

100

200

300

400

500

600

view, are unlikely to adapt to a new pricing regime if it is implemented directly through “prices to consumers,” particularly for a staple such as water.

Stormwater systems a niche play. Stormwater systems are typically designed to harvest rainwater in urban areas to supplement the water table, as well as treating the water early to reduce health risks.

Agriculture: More crop per drop?

In the United States, crop production uses almost 25% of the fresh water drawn from rivers, lakes, and aquifers each year, and 65%–70% of the freshwater that is not used for thermoelectric power. Groundwater represents an increasing share of total irrigation water in the United States, at roughly 42% in 2000 vs. 23% in 1950—and in absolute terms, the amount of groundwater used for irrigation has tripled. Most of this increase occurred by 1980, and since then demand has been roughly flat, albeit with some increases due to drought. Outside the United States, agriculture represents an even larger demand on freshwater supplies, at roughly 70% globally (and north of 80% in Asia and Africa). The highest volumes of water demand, globally, go to cultivate rice (21%), wheat (12%), corn (9%), and soybeans (4%).

EXHIBIT 23: TOTAL WATER USE (BN GAL/DAY) AND FOR IRRIGATION: STILL BELOW 1980 HIGHS

0

100

200

300

400

500

600

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 20000102030405060708090

Total withdrawals (LHS) Irrigation (RHS)

0

100

200

300

400

500

600

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 20000102030405060708090

Total withdrawals (LHS) Irrigation (RHS)Source: GAO

Over the past 50 years, crop yields have risen roughly 2.5x, with irrigation water requirements rising 2.2x. By the late 1990s, irrigation accounted for almost 40% of food production, even though it is used on less than 18% of the cultivated land. Where companies such as Monsanto and DuPont (DD, $33.76, NC) have focused on improving yields, we believe that after 2010 drought-resistant strains of cotton, soybeans, corn, and eventually rice could prove essential for gaining market share and premium prices on seeds. Drought resistance and water-efficiency are particularly important because irrigation systems can lose more than 60% of the water to evaporation and runoff (typically carrying soil nutrients and fertilizer with the water). In the near term, government subsidies (estimated at more than $2.5bn/year in the U.S. alone) reduce the cost of irrigation (to roughly $100/acre in the U.S.), which reduces the incentive for efficient resource use. Longer term, however, water stress could lead to a shift in government incentives, leading to a greater value for water-efficient crops (e.g., shorter life cycles to achieve similar yields).

EXHIBIT 24: IMPROVED DIETS ENTAIL MORE WATER CONSUMPTION

Food Water equivalent (m3/kg) Beef 15Lamb 10Poultry 6Cereals 1.5 Citrus Fruits 1Palm Oil 2

Source: FAO

With this context in mind, we favor the seed suppliers over the suppliers of irrigation equipment because seeds have faster turnover as consumables, and because irrigation equipment markets tend to be more focused and tilted towards marginal, more arid land.

Irrigation represents a significant perennial infrastructure requirement, as irrigation infrastructure can cost $1,000–$25,000/hectare depending on the design. Global irrigation investment is expected to run around $5bn/year for new




Page 135 of 212

irrigation projects, supplemented by another $5–$10bn/year for modernization of infrastructure built in the 1960s and 1970s, and a similar amount for on-farm projects. On a regional basis, irrigation projects have the most impact in the Middle East/Northern Africa (more than 53% of all water demand), followed by South Asia (36%, vs. 8% globally).

EXHIBIT 25: APPLICATION EFFICIENCY (%) OF VARIOUS ON-FARM IRRIGATION SYSTEMS

Irrigation System Application Efficiency (%) Earth canal 40-50 Lined canal 50-60 Pressure pipe 65-75 Hose irrigation 70-80 Sprinkler 75Microsprinklers 75-85 Drip irrigation 80-90

Source: World Bank

The most promising market, in our view, is micro- or drip irrigation. Studies have shown that well-designed systems can reduce water-use by 30%–70% and also increase crop yields by more than 20%. The use of microirrigation can also reduce the pace of salinization in irrigated land. California alone represents 72% of the demand for microirrigation, and California, Nebraska, Texas, Arkansas, and Idaho represent more than half the total demand for irrigation via sprinklers and microirrigation systems. Globally, drip irrigation represents less than 2% of the total market. The main hurdle has been cost: micro-irrigation and drip irrigation systems can cost 2x–3x as much to install and maintain as more conventional pipes or sprinklers.

EXHIBIT 26: IRRIGATED AREA (‘000 HECTARES, 1961–2003)

Region 1961 1965 1970 1975 1980 1985 1990 1995 2000 2003Africa/M.E. 7,389 7,770 8,434 8,943 9,315 10,073 11,000 12,463 13,162 13,370

Asia 95,523 96,713 109,37

3121,15

1131,65

9 141,780 155,52

5 181,767 193,169 193,89

0Europe 12,228 13,361 15,023 18,504 21,359 23,894 25,734 26,150 25,341 25,208 North America 17,950 19,526 20,939 22,822 27,571 27,454 28,893 30,465 31,426 31,264 Oceania 1,081 1,370 1,590 1,622 1,686 1,959 2,118 2,695 2,682 2,844 South America 4,661 5,070 5,673 6,403 7,382 8,269 9,494 10,155 10,488 10,522

Total 138,83

2 149,750 167,69

2188,14

5209,29

2 225,242 245,24

4 263,695 276,268 277,09

8Source: The World's Water 2006-2007, Peter H. Gleick, & Jefferies & Company, Inc. estimates

EXHIBIT 27: IRRIGATED AREA (‘000 HECTARES, 1961–2003)

0

50,000

100,000

150,000

200,000

250,000

300,000

1961 1965 1970 1975 1980 1985 1990 1995 2000 2003

Africa/Middle East/Oceania Asia EuropeNorth & Central Americas Oceania South AmericaFormer Soviet Union

Source: The World's Water 2006-2007, Peter H. Gleick, & Jefferies & Company, Inc. estimates




Page 136 of 212

Industrial water treatment: Energy efficiency, process optimization & conservation

Global water demand for industrial applications has been estimated at roughly 10% of global water consumption, or roughly 716bn m3/year. Treating this water represents a roughly $19bn market, with companies offering chemicals and filters for organic and inorganic contaminants and process solutions to optimize water efficiency for specific industrial processes. Water filtration, separation, and purification for industrial companies represent an adjacent $6.3bn market (out of a global filtration market of roughly $35bn).

The industrial water treatment market has been in vogue lately, mostly due to powerful global trends of water scarcity, tightening water quality standards, and increasing incentives to pass rising water treatment costs through to manufacturers. There is a clear tendency for water treatment activity, for example, to rise as a population industrializes. The industries that use the most water include paper, chemicals, and petrochemical refining, and primary metals processing.

Unlike the water utility market, where consumer behaviour is largely oblivious to cost dynamics, industrial end users face persistent competition from rival technologies and process innovations. As a result, the growth rate of water use in industry in the developed world has slowed in recent years as growth in production levels has been offset in part by efficiency gains. In the United States, for example, industrial water use was 11% lower in 2000 than in 1995. Indeed, in North America, which represents roughly 45% of the global market for industrial water treatment, volume growth has slowed to roughly 2%, down from 5%–7% in the late 1980s and early 1990s. In Western Europe (20%), growth is expected to remain around 3%–4%, while Japan (1%) has slowed to 1% or less. The industry consequently has focused more sharply on the markets in the rest of Asia and Latin America, which are expected to continue to grow 8%–15% as wastewater regulations converge on U.S. and European standards.

EXHIBIT 28: WATER TREATMENT DEMAND ($/CAPITA) VS. MANUFACTURING VALUE ADDED ($/CAPITA)

1,000 1,500 2,000 2,500 3,000 3,500MVA ($/person)

Wat

erTr

eatm

ent(

$/pe

rson

)

0

5

10

15

20

25

30

35

AustriaTaiwan

South KoreaBelgium

GermanySweden

JapanUnited KingdomFrance

ItalyNetherlands

Spain

Czech RepublicAustralia

HungaryRussia

Mexico

ArgentinaIndonesia

ChinaBrazil

Scatterplot of Water Treatment ($/person) vs. MVA ($/person)

United States

1,000 1,500 2,000 2,500 3,000 3,500MVA ($/person)

Wat

erTr

eatm

ent(

$/pe

rson

)

0

5

10

15

20

25

30

35

AustriaTaiwan

South KoreaBelgium

GermanySweden

JapanUnited KingdomFrance

ItalyNetherlands

Spain

Czech RepublicAustralia

HungaryRussia

Mexico

ArgentinaIndonesia

ChinaBrazil

Scatterplot of Water Treatment ($/person) vs. MVA ($/person)

United States

Source: Nalco

EXHIBIT 29: U.S. WATER TREATMENT VOLUME HAS SLOWED AS END MARKETS MATURED

Years Volume CAGR (%) 1986-1989 4.5% 1989-1992 6.4% 1992-1995 5.2% 1995-2000 3.9% 2001-2006 2.2%

Source: SRI

The industrial water treatment industry has also been instrumental in helping manufacturers introduce and optimize closed-loop processes that significantly reduce the amount of freshwater used. In the paper industry, for example,




Page 137 of 212

companies have reduced the amount of water used to produce paper by more than 50% over the past three decades.

The water treatment market distinguishes between non-chemical technologies, such as membrane technologies, UV disinfection, and ozone disinfection, and water treatment chemicals, particularly coagulants, flocculants, corrosion, and scale inhibitors, and biocides. Ion exchange resins and activated carbon together represent less than 15% of the total water treatment market.

Water treatment also represents an opportunity for the industrial gas suppliers, mostly through the supply of oxygen for ozone applications. We estimate that to treat drinking water one requires roughly 0.0011 scf of oxygen for every litre of water. For contrast, a world-scale oxygen plant might produce 10,000–102,000 scf/day. The demand for oxygen increases depending on the quality of the water, particularly its biological oxygen demand level for aeration treatments. For municipal wastewater (300–500 mg/l BOD), oxygen demand might run around 0.00133–0.0222 scf/litre. Industrial wastewater might have BOD levels as high as 5,000 mg/l, requiring roughly 0.2 scf/litre.

EXHIBIT 30: U.S. WATER RECYCLING RATES IN MAJOR INDUSTRIES (% OF TOTAL WATER CONSUMPTION)

Year Paper Chemicals Petroleum and Coal Primary Metals Manufacturing 1954 2.4% 1.6% 3.3% 1.3% 1.8% 1964 2.7% 2.0% 4.4% 1.5% 2.1% 1973 3.4% 2.7% 6.4% 1.8% 2.9% 1978 5.3% 2.9% 7.0% 1.9% 3.4% 1985 6.6% 13.2% 18.3% 6.0% 8.6% 2000 11.8% 28.0% 32.7% 12.3% 17.1%

Source: ITT

Energy

Single largest driver of water demand: The largest single use of water by industry is in the cooling process related to thermal power generation, including both treatment of raw water for cooling (which prevents bacterial infection and equipment degradation) along with treatment of water consumed in treating flue gas in thermal plants.

Thermo-electric power alone accounts for roughly 52% of freshwater use and 96% of saline-water use. In the United States, thermoelectric plants use an estimated 270bn litres/day of seawater and another 607.5bn litres/day of freshwater. Most of the freshwater (roughly 594bn litres/day) is returned to the source, leaving roughly 15.75bn litres/day lost through evaporation. With environmental considerations encouraging more use of closed-loop cooling systems at thermo-electric facilities, the DoE estimates that freshwater demand by these facilities could more than double by 2030 as new plants are brought onstream and older open-loop facilities are retrofitted.

The key cyclical driver of demand for water treatment is overall power industry capital expenditure, with support from secular trends such as the increased focus on “clean coal” plants globally, which demand greater water resources for flue gas scrubbing. New refining capacity also contributes to water demand, as refineries use an estimated 4.5-11.3 litres of water for every four litres of product produced.

EXHIBIT 31: WATER INTENSITY FOR POWER GENERATION (GAL/MWH)

Steam Condensing Other Energy source Withdrawal Consumption Withdrawal Consumption Coal mining/slurry 110-230 35-145 Thermo-electric

Open Loop Cooling 20,000-50,000 300 30 30 Closed Loop Cooling 300-600 300-480 30 30Dry Cooling 0 0 30 30

Nuclear Mining 45-150 Open Loop Cooling 25,000-60,000 400 30 30 Closed Loop Cooling 500-1,100 400-720 30 30

Geothermal 2,000 1,400 Solar tower 750-920 750-920 8 8Natural Gas Combined Cycle

Open Loop Cooling 7,500-20,000 100 21 21Closed Loop Cooling 230 180 21 21




Page 138 of 212

Coal IGCC 250 200 140 140Hydroelectric 4,500

Source: Department of Energy Water intensity has declined: The energy market has been one of the primary beneficiaries of new technologies and processes to improve water efficiency (albeit mostly due to regulatory pressure). As a result, energy production now takes an average 94.5 l/kWh, down from 283.5 l/kWh in 1950. Exporting these processes and designs to rapidly industrializing emerging markets represents, in our view, one of the most sustainable long-term opportunities for the water treatment and process control companies.

EXHIBIT 32: WATER CONSUMPTION BY ENERGY SOURCE: ALTERNATE FUELS

Application Litres/Day or /product Litres/MMBTU

Refining 4.5-9bn

4.5-11.3 l/l of product 31.5-81

Natural gas processing & pipelines 1.8-2.3bn 9-13.5

Coal mining & transportation via slurry 315-1,125m 45-450 l/ton

4.5-27 49.5-108 for slurry

Coal gasification 18 t/t of methanol produced 49.5-117 Oil shale 9-22.5 l/l of product 67.5-171 Oil sands 18-22.5 l/l of product 90-337.5 Enhanced Oil Recovery up to 1,575 l/l extracted 67.5-4,500

Oil & gas extraction 220.5m 11.3-13.5 l/boe -

Synfuels 20.7-30.6 l/l of product 184.5-270

Uranium - 4.5-27.0 for mining 31.5-36 for processing and enrichment

Ethanol 21.2 l/l of product (dry mill) 58.5-652.5, or 11,250-130,500 (including crop)

Biodiesel 4.5 litre/litre produced 18.9, or 63,000-337,500 including crop

Hydrogen - 22.1 via natural gas 94.5-990 via electrolysis

Source: Department of Energy, API, Nalco

Alternative energy could exacerbate water requirements: Alternative energy, moreover, has tended to be more water intensive, creating an opportunity for water treatment and management firms such as Buy-rated Nalco. Water consumption could also create a bottleneck for certain technologies, particularly coal gasification and coal-to-olefins, where the feedstocks are typically available in arid climates, creating logistical complications for potential energy or chemical facilities. The Canadian tar sands, for example, use roughly 18-22.5 litres of water for every four litres of product produced, while coal-to-methanol processes can require as much as 18x the volume of water as the ethanol produced. The renewable energy technologies that represent the least potential drain on freshwater supplies include wind power, photovoltaics, air-cooled geothermal systems, tidal energy systems, and hydroelectric capacity that do not disrupt river flows.

Implications for energy demand. Almost 4% of U.S. power generation is used for water supply and treatment, comparable to the paper industry, and roughly 60% of the total energy requirements for the chemical industry. Energy can represent as much as three-quarters of the cost of sourcing water. The most significant driver of energy costs is the need to pump freshwater from aquifers: pumping 4.5 million litres from 120 feet can require 540kWh, and 4x that for extracting water from 400 feet. In the near-term, droughts can reduce hydroelectric power production, forcing energy-intensive industries to place more reliance on natural gas for peaking capability, particularly in the summertime.

EXHIBIT 33: ENERGY REQUIREMENTS IN CALIFORNIA FOR WATER SUPPLY AND TREATMENT (KWH/M GALLONS)

Water Segment Low High Supply & Conveyance 0 16,000 Treatment 100 1,500 Distribution 700 1,200 Wastewater Collection & Treatment 1,100 4,600 Wastewater Discharge 0 400Total 1,900 23,700

Recycled Water Treatment & Distribution For Non-Potable Uses 400 1200




Page 139 of 212

Source: DoE, CEC

Biofuels

The water consumption entailed by biofuel production has emerged as a particularly active source of controversy, but also represents an opportunity for providers of water treatment and conservation technologies. We estimate the addressable market opportunity at north of $250m in the United States alone.

With current technology, a 450m l/year ethanol plant requires as much water as a town of 5,000 people, mostly due to cooling towers and evaporators. Current estimates are that ethanol production requires eighteen litres of water for every four litres of ethanol produced, and as much as 32 litres for the older, more inefficient plants. Petrochemical refineries require less than half as much, partly due to more sophisticated process designs. At a typical ethanol plant, roughly 60% of the water consumption occurs at the cooling tower, and 35% is lost as water vapor as part of the drying process used to generate DDGS as a co-product. Cellulosic ethanol is expected to consume nine to 27 litres/litre of fuel, with thermochemical processes at the lower end of the range. Biodiesel production, in contrast, is expected to consume nine to 14 litres/litre—and only four litres of freshwater for every four litres of fuel (and better energy intensity than ethanol as well). These values compare with the average US consumption of roughly 810 litres/person/day.

EXHIBIT 34: U.S. BIOFUELS PRODUCTION, 2006

Fuel Feedstock U.S. Production in 2006 Ethanol Corn 22.1bn litres

Sorghum <450m litres Cane sugar 2,700m litres imported from Brazil and the Caribbean

Cellulose Demonstration facilities in Canada, Asia Biodiesel Soybean oil 405m litres

Other vegetable oils <45m litres Recycled grease <45m litres

Cellulose None Source: CRS

These water figures, however, are only for the biofuel facilities themselves. The implied water requirements increase when the feedstocks are included. For example, corn in Nebraska can require 9,450 litres/bushel from irrigation. This translates into another 3,510 litres/litre of ethanol, or 195x the water consumption at the ethanol facility itself. Importantly, the water required to grow a crop can vary widely depending on the prevailing climate and soil quality. In the United States, for example, corn requires less water than soybeans and cotton in the Pacific and Mountain regions, and more in the Plains.

EXHIBIT 35: AVERAGE WATER USE (GAL/BUSHEL) FOR CORN IN THE UNITED STATES, 2003

Source: USDA




Page 140 of 212

Average Gal. Per Bushel501 – 1,0001,001 – 2,0002,001 – 3,0003,001 – 4,0004,001 – 5,905No Data

Average Gal. Per Bushel501 – 1,0001,001 – 2,0002,001 – 3,0003,001 – 4,0004,001 – 5,905No Data

501 – 1,0001,001 – 2,0002,001 – 3,0003,001 – 4,0004,001 – 5,905No Data

Some studies have attempted to evaluate the aggregate impact that biofuels could have on global water consumption. A Sri Lankan study, for example, estimated that by 2030 biofuels production, based on industry targets, could represent as much as 3% of global freshwater use for agriculture and 11%–17% of crop water use in the EU, Brazil, and the United States (Exhibit 36). In the United States, biofuel production could eventually reach 20% of total irrigation spending, which could place a considerable strain on local aquifers. Biofuels could be a significant incremental burden for China and India, where irrigation accounts for a significant amount of current food production, with more than three-quarters of grain, vegetable, and cotton production coming from irrigated areas.

EXHIBIT 36: ETHANOL LAND AND WATER USE (2005)

Country % of global ethanol

production % of total crop area % of crop water* % of irrigation Brazil 41.0% 5.0% 10.7% 3.5% USA 35.1% 3.5% 4.0% 2.7% China 9.9% 1.1% 1.5% 2.2% India 4.8% 0.2% 0.5% 1.2% France 2.3% 1.2% 1.8% South Africa 1.1% 1.1% 2.8% 9.8% UK 1.1% 2.4% 2.5%

Global ethanol 0.8% 1.4% 2.0% Source: IWMI, Jefferies & Company, Inc. estimates * Share of water evapotranspirated by crops per year. Does not include farm run-off.

That significant and growing draw on water resources creates an opportunity for the water sector, as a range of technologies can help mitigate these issues. Closed loop processes should dramatically lower water consumption while precision agriculture can tailor water consumption at the crop level to micro variations in soil quality, enhancing yields while delivering net reductions in total water consumption.

Food and Beverage: Bottled water a runaway success

Bottled water represents a high-profile market, with an estimated $95–$98bn in sales in 2008, growing in the high single digits. In the United States, bottled water demand is ahead of beverages such as alcohol and coffee. Successful branding strategies have led to bottled water prices running more than hundreds of times that of urban tapwater, as well as roughly double gasoline prices at the pump (on a per liter basis). The popularity of bottled water as a beverage, coupled with strict FDA rules as to what can be labeled “spring water”, has led to incentives for bottlers to target upstream water supplies, a source of increasing environmental controversy.

EXHIBIT 37: GLOBAL BOTTLED WATER DEMAND, 1997–2004 (‘000 M3)

Region 1997 1998 1999 2000 2001 2002 2003 2004Europe 34,328 36,074 39,965 42,276 44,520 47,037 51,768 53,661 North America 25,398 25,822 29,695 31,850 34,734 38,349 41,778 44,715 Asia 12,472 14,820 17,647 21,170 24,824 29,783 32,795 35,977 South America 5,484 6,362 7,323 8,528 9,915 11,437 12,677 13,607 Africa/Middle East 2,459 2,808 3,092 3,456 3,837 4,302 4,499 4,823 All Others 508 1,953 737 891 1,033 1,592 1,407 1,597 Total 80,649 87,839 98,459 108,171 118,863 132,500 144,924 154,380

Source: The World's Water 2006-2007, Peter H. Gleick p. 281

The food and beverage industry also uses water as an ingredient for other products, of course, and as a medium for processing — notably as a cleaning agent, for boiling, or for conditioning of raw materials. In the United States, the beverage industry uses roughly 58.5bn litres/year, or 8%–10% of total industrial demand. Key drivers including tighter regulatory requirements (integrated pollution prevention & control), initiatives to improve process efficiency and reduce cost (e.g., water recycling), and the introduction of “best practices” in the emerging markets. While this is a difficult sector to find direct investment plays, indirect beneficiaries of moves towards greater water efficiency include Buy-rated Christ Water and Hold-rated Ecolab in the United States, and companies that provide water metering. Christ Water, which designs turkey systems for food and beverage players, is directly benefiting from increased regulation in the sector (necessitating retrofit or entirely new equipment) and also increased focus on efficiency in new installation. Ecolab should benefit as the water utilities increasingly shift the onus for water treatment back to the sources of the wastewater, even the small restaurants and hospitals.




Page 141 of 212

Pharmaceuticals: A targeted play

Water is the largest volume element used in the pharmaceutical production process. Key uses of water in the industry include as an ingredient or by-product in the production process and as a cleaning agent for vessels, equipment, and packing materials. Each process requires different grades of water quality and at the same time different preparations of medicines demand different levels of water purity—an opportunity for companies that provide filtration, biocides, and other consumables. Along with the requirement for pure water to avoid potential introduction of bacteria into end products, another key element of water management in the pharmaceutical sector is the treatment of wastewater. Environmental concerns over pollutant discharges from the pharmaceutical industry and limitations on discharges into public wastewater treatment facilities have necessitated the creation of in-house water treatment facilities.

Desalination: Rising energy costs threaten economics

While desalination plants have historically been a niche market due to the high relative cost of production, improved technology has driven down costs and made the process comparable in price to traditional sources of potable water. While the technology varies, the underlying principle is to use energy to separate freshwater from saltwater. The World Business Council for Sustainable Development has estimated that the cost of freshwater delivered using desalination has fallen almost 80% over the past 15 years—and in some regions appears economical. Studies in the early-to-mid-1990s estimated the cost of desalination at $1,000–$2,200/acre-foot, or $3–$7/4,500 litres. While technology-related process costs have come down sharply in recent years (by more than half), energy costs are significant, as desalination can require 2,500–30,000 kWh/acre-foot depending on the salt concentration in the water being processed. Large-scale desalination plants (over 45m litres/day) can process seawater for $1.70–$4.00/4,500 litres, whereas plants processing brackish water can deliver freshwater for $0.50–$4/4,500 litres. For example, the $250m 100m m3/year Ashkelon desalination plant in Israel started production in 2006 with an estimated cost of $0.53/m3. Rising energy costs, however, have likely led to a sharp increase since then.

EXHIBIT 38: SOME ESTIMATES FOR DESALINATION COSTS DEPENDING ON WATER SUPPLY (ESTIMATES FROM 1996–2003)

Brackish Water Seawater Fixed costs (incl. 20-year depreciation) 54% 37% Electric power 11% 44%Labor 9% 4% Membrane Replacement 7% 5%Maintenance 9% 7% Consumables (Chemicals) 10% 3%Cost ($/m3) $0.25-$0.70 $0.45-$6.56

Source: Tamim Yuounos, “The Economics of Desalination”, Journal of Contemporary Water Research & Education, Dec. 2005

Desalination has been most popular in arid countries, particularly in the Middle East and North Africa. Globally, desalination accounts for approximately 3bn m3/year, with most of the production occurring in Kazakhstan (43%), Saudi Arabia (23%), the United Arab Emirates (13%), Kuwait (7%), and Qatar (3%). In the United States, the most interest arises in states such as Arizona, California, Florida, New Mexico, and Texas.

EXHIBIT 39: KEY DESALINATION METRICS

Desalination Method Thermal Separation Reverse Osmosis Electro Dialysis

Operating Temperature (Deg. Celsius) 35-120 0-40 0-65

Salt content (mg/l) 30,000-500,000 500-50,000 500-3,000 # of units (2003) 2,925 9,180 1,495

% 21.5% 67.5% 11.0% Capacity

(m acre-feet/year, or bn of litres/day) 16.2 13.5 8.1

% 51.0% 43.1% 5.9% Energy used

(kWh/'000 litres) 67.5 49.4-63

Other comments Energy needs down from 20-25

kWh/4,500 litres in 1990

Capital intensity: $2-$3/4,500 litres for a 112.5m litres/day plant, vs. $4-$5 for a 45m litres/day plant.

Projects in the late 1980s cost $6-$12/4,500 litres. 90% reduction in membrane costs since 1990

Source: Company information, Jefferies & Company, Inc. estimates




Page 142 of 212

Looking ahead, desalination should remain one of the more dynamic and faster changing sub segments of the water industry. Vastly increased size (with the newest plants more than 7x as large as the plants built in the 1980s and 1990s) and new technologies (particularly using reverse osmosis membranes) have brought the delivery cost down to levels comparable to traditional water sources in many areas. While many argue that the easy gains have been made based on existing technology, GE for one is now targeting $0.10/m3 in the not too distant future. That level would be game changing for the entire industry and could lead to incremental market share shifts over the next few years.

EXHIBIT 40: NUMBER OF DESALINATION PLANTS BUILT (RHS) AND AVERAGE CAPACITY OF NEW PLANTS (M3/DAY, LHS)

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

1945

1948

1951

1954

1957

1960

1963

1966

1969

1972

1975

1978

1981

1984

1987

1990

1993

1996

1999

2002

0

50

100

150

200

250

300

350

400

450

500

Number of Plants (RHS) Average Capacity per Plant (LHS)

Source: Jefferies & Company, Inc. estimates, “The World's Water 2006-2007”, Peter H. Gleick

EXHIBIT 41: DESALINATION CAPACITY GROWTH (% YOY, LHS) AND CUMULATIVE CAPACITY (M M3/DAY, RHS)

0%

10%

20%

30%

40%

50%

60%

70%

80%

1950

1952

1954

1956

1958

1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

0

5

10

15

20

25

30

35

40

% YoY (LHS) Cum. Capacity (m m3/d, RHS)

Source: Jefferies & Company, Inc. estimates, “The World's Water 2006-2007”, Peter H. Gleick




Page 143 of 212

6. How do water stocks trade?

Separating the Baby from the Bath Water

Likely to nobody’s surprise, relative valuations in the water sector tend to be driven by end market growth dynamics, barriers to entry/sustainability of business models, and also technological complexity. The result is that the filtration and metering players tend to trade at the highest multiples, while more plain vanilla industrial plays are hardly accorded a water premium. US utilities are a notable exception to that analysis due to a favourable regulatory environment and aggressive consolidation which has left the remaining players trading on a scarcity premium.

EXHIBIT 42: WATER SUBSECTOR MULTIPLES

Ut ilit ies - U.S.M etering

Filt rat ionUt ilit ies - Europe

Diversif ied

Flow Technology

Automat ion

Treatment

0x

5x

10x

15x

20x

25x

30x

Increasing Complexit y and Growth Prospects

Source: Bloomberg, Jefferies & Company, Inc.

While relative subsector valuations have largely held the pattern outlined above, there has been a great deal of volatility over the past 15 years.

• Filtration multiples, while generally at a premium to the overall sector, have been very volatile, descending as low as 60% of the broader market in the late 1990s and rising to a 30% premium recently.

• The Flow Technology plays have traditionally traded at a relative discount to the other subsectors in the space, reflecting in part, the lower average exposure to water in the business mix. Not surprisingly given their more diverse business mix, longer-term trends in relative valuations have coincided with the longer-term upcycles in the industrial economy.

• Solutions & Treatment valuations have tended to remain in the 15x–25x range, on average, with a brief period of elevated valuations due to depressed earnings at a small set of companies, including Calgon Carbon and Nalco.

• US Water Utility forward P/E multiples expanded from the 13x–17x range in the late 1990s to the 20x–25x range this decade. This reflects both a favorable regulatory environment and, more importantly, ongoing consolidation in the sector and a consequent scarcity premium for the remaining players.

• The trend in US utility trading ranges is notable in its divergence from the experience of the EU utilities. While pan-European utility M&A remains in its infancy, the space has seen significantly more privatization and consolidation than the United States. As a result of the relative maturity of the space and no shortage of investment vehicles, a lid has been kept on multiple expansion.

• The Conglomerates, given their broad industry exposures, have tended to post valuation multiples generally in line with the S&P 500.


Forw

ard

P/E



Page 144 of 212

EXHIBIT 43: HISTORICAL TRADING RANGES FOR WATER SUBSECTORS

10x

14x

18x

22x

26x

30xFlo w Techno lo gy TreatmentFiltratio n A uto matio nS&P

5x

10x

15x

20x

25x

30x

U.S. Utilit iesE.U. Utilit iesS&P

Source: Bloomberg, Jefferies & Company, Inc.

Valuations matter

Along with divergence across subsectors, another key trading relationship has been a clear value bias in the sector.

While individual companies can support elevated multiples for extended periods, their premium to the sector can befairly stable. As a result, waves of investor optimism or positive developments tend to ripple through the group fairlyquickly, so there is significantly less downside risk to buying the less expensive companies in the sector. Indeed,when we look at trading baskets of water shares based on forward P/E ratios, we find a clear bias in favor ofcheaper shares.

• Stocks with high multiples are likely to bump into a ceiling, while stocks with low multiples have plenty of room toimprove. Given the scarcity of compelling stories in the sector, coupled with the tendency for company consensusestimates to move in tandem, any irrational surges in a company’s shares tends to correct within a couple ofmonths.

• Investing each month in the five companies with the lowest NTM P/E multiples would have turned $1 into $10.63over the past 10 years (a 26.7% CAGR), versus only $0.92 (an average annual loss of 0.9%) from purchasing thecompanies with the highest multiples. By contrast, the S&P 500 turned $1 into $1.53 in the same period.

• Even simply investing in those companies with below average NTM P/E multiples outperformed selecting theirmore expensive peers by 1,200bps a year ($5.24 vs. $1.72).


Forw

ard

P/E

Forw

ard

P/E

Jan-

95

Jan-

99

Jan-

97

Jan-

01

Jan-

03

Jan-

05

Jan-

07

Jan-

95

Jan-

99

Jan-

97

Jan-

01

Jan-

03

Jan-

05

Jan-

07



Page 145 of 212

EXHIBIT 44: PERFORMANCE OF WATER STOCK PORTFOLIOS WHEN SELECTED MONTHLY BASED ON NTM P/E, 1991–2007 (EX-DIVIDENDS)

Lowest 5 Highest 5 Below Avg. Above Avg. 1990 (17.3%) 16.7% (9.7%) 14.4% 1991 36.0% 43.1% 28.7% 46.2% 1992 26.9% 14.8% 15.8% 6.4% 1993 33.3% 1.7% 25.3% 7.8% 1994 (11.0%) 0.1% (3.4%) 4.5% 1995 18.5% 28.3% 18.9% 32.2% 1996 31.1% (5.0%) 21.4% 15.5% 1997 55.7% 47.4% 33.7% 26.6% 1998 (23.0%) (8.4%) (7.2%) 4.3% 1999 30.8% 7.8% 19.6% 7.1% 2000 5.1% (33.0%) 12.2% 2.4% 2001 81.5% 13.9% 44.4% 4.0% 2002 (29.0%) (26.2%) (8.7%) (19.4%) 2003 17.4% 38.8% 32.9% 28.1% 2004 59.8% 24.0% 30.3% 15.0% 2005 143.3% (17.6%) 29.7% 0.1% 2006 22.0% 5.0% 32.1% 10.2% 2007 39.9% 10.6% 7.6% 10.1% 2008 (Year-to-date) (11.3%) (2.6%) (1.8%) (6.9%) 10-year CAGR 26.7% -0.9% 18.0% 5.6% $1 becomes (10-yr) $10.63 $0.92 $5.24 $1.72 $1 becomes (since 1991) $36.74 $1.54 $14.26 $3.93


Historical Performance

Including dividends, the water sector has generated a 14.7% CAGR over the past 15 years, versus 14.2% for the S&P 500. The bulk of the sector’s outperformance has occurred since the late 1990s. Since 1998, for example, the sector has generated a 7.8% CAGR vs. 2.9% for the S&P 500. Since 2003, the sector has delivered a 19.5% CAGR vs. 10.2% for the S&P 500.

EXHIBIT 45: WATER STOCKS: ABSOLUTE PERFORMANCE, 1985–2008


Since 1985, the water solutions and treatment companies have provided the strongest returns within the sector, rallying almost 14x.

0

200

400

600

800

1000

1200

1400

1600

Jan-

85

Jan-

86

Jan-

87

Jan-

88

Jan-

89

Jan-

90

Jan-

91

Jan-

92

Jan-

93

Jan-

94

Jan-

95

Jan-

96

Jan-

97

Jan-

98

Jan-

99

Jan-

00

Jan-

01

Jan-

02

Jan-

03

Jan-

04

Jan-

05

Filtration Flow Tech Solutions & Treatment Utilities




Page 146 of 212

US water utilities lagged the S&P 500 in the 1980s and early 1990s, but relative performance has improved in recent years. Water utilities have provided steadier, less volatile returns, with a 7.1% CAGR over the past 10 years and 12.1% over the past five years. Dividends have made a significant contribution to returns: including dividends, the water utilities have delivered a 10.1% CAGR over the past 10 years and 15.7% over the past five years. In absolute terms, the group was basically flat from 1985 through 1995, with consistently solid results since then due to both declining interest rates and a favorable regulatory environment. Since 2003, the utilities have traded basically in line with the S&P 500.

EXHIBIT 46: LONG-TERM RETURNS, INCLUDING DIVIDENDS, FOR WATER STOCKS

Segment 5 year 10 year 15 year Filtration 17.7% 8.9% 16.9% Conglomerates 6.0% 7.1% 8.6% Flow Technologies 26.8% 8.9% 16.7% Water Rights 27.5% -13.8% -4.2% Solutions & Treatment 17.2% 7.0% 14.7% Utilities 15.7% 10.1% 17.5% Water Stock Average 19.5% 7.8% 14.7% S&P 500 10.2% 2.9% 14.2%


EXHIBIT 47: LONG-TERM RETURNS, INCLUDING DIVIDENDS, FOR WATER STOCKS, RANKED BY 10-YEAR CAGR

10.2%

17.2%

6.0%

19.5% 17.7%

26.8%

15.7%

2.9%7.0% 7.1% 7.8% 8.9% 8.9% 10.1%

0%

5%

10%

15%

20%

25%

30%

S&P 500 Solutions &Treatment

Conglomerates WaterAverage

Filtration FlowTechnologies

Utilities

5 year 10 year

10.2%

17.2%

6.0%

19.5% 17.7%

26.8%

15.7%

2.9%7.0% 7.1% 7.8% 8.9% 8.9% 10.1%

0%

5%

10%

15%

20%

25%

30%

S&P 500 Solutions &Treatment

Conglomerates WaterAverage

Filtration FlowTechnologies

Utilities

5 year 10 yearSource: Capital IQ, Jefferies & Company, Inc. estimates




Page 147 of 212

7. What about consolidation?

Along with the secular trend of increased demand for water and equipment to improve water efficiency, another key industry trend has been that of consolidation. The most visible trend has been consolidation in US water utilities, as more than 80% of the utilities serve fewer than 10,000 people, too small a base for the infrastructure investments that are required over the next decade.

Global industrial players such as 3M, Siemens, and GE, have also driven consolidation among the technology-focused solution providers (e.g., Zenon, Trojan, Ionics, and Cuno), which have been acquired for multiples as high as 16x-20x EBITDA due to their attractive long-term growth prospects. This has left only a handful of publicly traded companies focused purely on water technology. The sector has also seen acquisitions, albeit at less elevated multiples, across the remainder of the value chain from chemical and equipment manufacturers that are seeking to integrate with engineering businesses.

Looking towards the next three to five years, we expect sustained M&A activity in the utility space as consolidation remains in its early stages, and also in the $19bn water treatment market, which remains fairly fragmented with Nalco (17% share) as the only pure-play. Longer term, we expect emerging markets players to acquire US rivals to establish global positions, particularly in the more commoditized parts of the value chain (e.g., valves and pumps).

EXHIBIT 48: WATER INDUSTRY M&A MULTIPLES, 1996–PRESENT

0x

5x

10x

15x

20x

25x

30x

35x

40x

Jan-96 Jan-98 Jan-00 Jan-02 Jan-04 Jan-06 Jan-08

Zeno n

Tro jan

Specialized Filtratio nTechno lo gy

Jacuzzi

U.S. F ilterU.S. Filter

Nat'lWaterwo rks

Thames Water

Saur

0.0x

0.5x

1.0x

1.5x

2.0x

2.5x

3.0x

3.5x

4.0x

4.5x

Jan-96 Jan-98 Jan-00 Jan-02 Jan-04 Jan-06 Jan-08

Everpure

Io nics

Eco lo chem Zeno n

Source: Jefferies Intl., Company Documents


EV/E

BIT

DA

EV/S

ales



Page 148 of 212

Laur

ence

Ale

xand

er, C

FA

, lal

exan

der@

Jeffe

ries.

com

, (21

2) 2

84-2

553

Appe

ndix

:Wat

erC

ompa

nyC

ompa

rison

sEX

HIB

IT49

:WAT

ERU

NIV

ERSE

2007

2008E

2009E

2007

2008E

2009E

2007

2008E

2009E

2007

2008E

2009E

Filtration/Treatment

BW

TA

GB

WT

AV

NC

15.1

268

0.7

0.7

0.6

6.7

7.2

6.5

8.4

9.2

9.0

10.2

10.6

10.3

2.26

CH

RIST

WA

TER

TEC

HN

OLO

GY

AG

CW

TA

VB

uy2.

140

0.3

0.3

0.2

10.2

7.4

5.4

17.5

13.3

8.4

15.8

10.0

0.21

CLA

RCO

RIN

CC

LCU

SN

C29

.91,

516

1.6

1.3

1.1

9.7

7.6

6.3

11.5

9.1

7.4

16.6

15.8

14.0

1.89

DO

NA

LDSO

NCO

INC

DCI

US

NC

30.2

2,37

11.

21.

00.

88.

67.

05.

710

.68.

56.

913

.912

.811

.2n/

aG

LVIN

C-A

GLV

/ACN

NC

4.9

117

0.3

0.3

0.2

16.1

4.7

3.0

n/a

7.6

4.5

n/a

10.4

6.3

0.22

HY

FLU

XLT

DH

YF

SPN

C1.

4375

14.

32.

21.

525

.69.

86.

133

.012

.17.

322

.711

.88.

10.

30K

UR

ITA

WA

TER

IND

UST

RIE

SLT

D63

70JP

NC

1860

247,

008

1.0

1.0

0.8

5.3

5.0

3.4

7.0

6.8

4.8

13.1

12.6

10.9

1.31

LYD

ALL

INC

LDL

US

NC

610

50.

3n/

an/

a3.

2n/

an/

a6.

2n/

an/

a11

.310

.211

.7n/

aM

ET-P

RO

CO

RPM

PRU

SN

C9

142

1.2

n/a

n/a

8.0

n/a

n/a

9.0

n/a

n/a

11.9

13.2

11.2

n/a

MIL

LIPO

RE

CO

RP

MIL

US

NC

54.5

3,00

52.

82.

21.

911

.98.

67.

319

.512

.010

.021

.615

.413

.50.

73O

RG

AN

OCO

RP

6368

JPN

C60

735

,176

0.5

n/a

n/a

7.2

n/a

n/a

9.6

n/a

n/a

14.2

15.2

11.5

1.12

PALL

COR

PPL

LU

SN

C27

.23,

248

1.4

1.2

n/a

7.4

6.3

n/a

9.2

8.0

n/a

15.4

12.3

11.1

n/a

POLY

POR

EIN

TERN

ATI

ON

AL

INC

PPO

US

NC

14.9

660

2.7

2.3

2.0

9.4

8.0

6.9

13.8

11.8

10.0

n/a

15.8

12.9

n/a

SHA

WG

RO

UP

INC

SGR

US

NC

15.9

1,32

60.

40.

20.

114

.73.

01.

620

.73.

51.

8n/

a7.

1n/

aSI

NO

MEM

TECH

NO

LOG

YLT

DSI

NO

SPN

C0.

1360

1.9

n/a

n/a

8.0

n/a

n/a

9.0

n/a

n/a

9.1

10.0

8.6

1.51

Average

1.4

1.2

0.9

10.1

6.8

5.2

13.2

9.3

7.0

14.6

12.2

10.9

1.06

U.S.Utilities

AM

ERIC

AN

STA

TES

WA

TER

CO

AW

RU

SN

C27

.547

42.

61.

5n/

a8.

34.

6n/

a12

.06.

5n/

a17

.016

.014

.31.

99A

QU

AA

MER

ICA

INC

WTR

US

NC

13.8

1,86

45.

1n/

an/

a10

.1n/

an/

a14

.2n/

an/

a19

.218

.916

.11.

51A

RTE

SIA

NR

ESO

UR

CES

CO

RP-

CL

AA

RTN

AU

SN

C14

.292

3.5

n/a

n/a

8.7

n/a

n/a

11.6

n/a

n/a

15.4

16.3

13.6

n/a

CA

LIFO

RNIA

WA

TER

SERV

ICE

GR

PC

WT

US

NC

28.7

594

2.4

n/a

n/a

9.2

n/a

n/a

14.2

n/a

n/a

19.1

17.1

14.4

1.42

CO

NN

ECTI

CUT

WA

TER

SVC

INC

CTW

SU

SN

C22

.018

54.

7n/

an/

a11

.4n/

an/

a16

.3n/

an/

a20

.717

.517

.0n/

aM

IDD

LESE

XW

ATE

RC

OM

SEX

US

NC

14.0

186

3.7

n/a

n/a

10.3

n/a

n/a

14.0

n/a

n/a

15.9

15.1

14.7

n/a

PEN

NIC

HU

CK

CO

RP

PNN

WU

SN

C20

.085

4.7

n/a

n/a

12.2

n/a

n/a

18.3

n/a

n/a

23.5

22.9

19.8

n/a

SJW

COR

PSJ

WU

SN

C21

.239

02.

9n/

an/

a9.

4n/

an/

a14

.4n/

an/

a20

.217

.415

.0n/

aSO

UTH

WES

TW

ATE

RCO

SWW

CU

SN

C8.

321

71.

7n/

an/

a11

.4n/

an/

a18

.0n/

an/

a-2

5.2

27.7

19.0

n/a

YO

RK

WA

TER

COY

OR

WU

SN

C11

.112

56.

3n/

an/

a11

.4n/

an/

a14

.0n/

an/

a19

.519

.516

.6n/

aAverage

3.8

1.5

n/a

10.2

4.6

n/a

14.7

6.5

n/a

14.5

18.8

16.1

1.64

EuropeanUtilities

RW

EA

GRW

EG

RN

C53

.027

,740

0.7

0.6

0.5

3.6

3.5

2.5

5.4

4.5

3.2

11.2

8.7

7.2

0.49

SEC

HE

ENV

IRO

NN

EMEN

TSC

HP

FPN

C42

.536

71.

71.

61.

06.

25.

73.

69.

78.

85.

615

.311

.19.

10.

49SE

VER

NTR

ENT

PLC

SVT

LNN

C10

702,

521

1.6

1.6

1.5

3.7

3.2

3.0

5.5

5.1

4.6

11.9

11.0

10.2

n/a

SUEZ

SASZ

EFP

NC

23.0

30,1

060.

90.

90.

76.

15.

74.

48.

47.

65.

87.

48.

57.

70.

73V

EOLI

AEN

VIR

ON

NEM

ENT

VIE

FPN

C22

.510

,632

0.8

0.7

0.6

6.6

6.0

4.7

10.1

10.2

7.9

10.4

10.8

9.6

0.99

Average

1.1

1.1

0.9

5.2

4.8

3.6

7.8

7.2

5.4

11.3

10.0

8.7

0.68

Company

Ticker

Price

MktCap

Rating

PEG

EV/Sales

EV/EBITDA

EV/EBIT

P/E

Sou

rce:

Blo

ombe

rg,J

effe

ries

&C

ompa

nyIn

c.es

timat

es

Alt

ern

ativ

e S

olu

tio

ns

- W

ater

Ple

ase

see

impo

rtan

t dis

clos

ure

info

rmat

ion

on p

ages

208

- 2

10 o

f thi

s re

port

.P

age

149

of 2

12

Laur

ence

Ale

xand

er, C

FA

, lal

exan

der@

Jeffe

ries.

com

, (21

2) 2

84-2

553

EXH

IBIT

49:W

ATER

UN

IVER

SE(C

ON

TIN

UED

)

2007

2008E

2009E

2007

2008E

2009E

2007

2008E

2009E

2007

2008E

2009E

LargeCap/DiversifiedWater

GEN

ERA

LEL

ECTR

ICC

OG

EU

SN

C19

.019

9,21

13.

82.

01.

818

.29.

28.

325

.514

.213

.08.

79.

610

.0n/

aIT

TCO

RP

ITT

US

NC

41.6

7,55

11.

00.

80.

67.

35.

84.

88.

96.

85.

510

.19.

98.

91.

53SI

EMEN

SA

G-R

EGSI

EG

RN

C47

.143

,096

0.7

0.7

0.5

6.1

6.6

4.0

10.9

10.3

5.6

11.1

9.5

7.3

0.41

Average

1.9

1.1

1.0

10.6

7.2

5.7

15.1

10.4

8.1

10.0

9.7

8.7

0.97

FlowTechnology

CIR

CO

RIN

TERN

ATI

ON

AL

INC

CIR

US

NC

32.9

553

0.8

n/a

n/a

7.3

n/a

n/a

9.0

n/a

n/a

n/a

8.3

7.2

0.22

CR

AN

EC

OC

RU

SN

C19

.61,

172

0.5

0.5

0.4

3.7

3.4

2.4

4.6

4.0

2.6

n/a

5.7

5.2

n/a

FLO

WSE

RVE

COR

PFL

SU

SN

C55

.93,

206

0.9

n/a

n/a

7.2

n/a

n/a

8.7

n/a

n/a

12.3

7.3

6.3

0.24

GO

RMA

N-R

UPP

CO

GR

CU

SN

C24

.641

11.

2n/

an/

a9.

4n/

an/

a11

.6n/

an/

a18

.014

.213

.2n/

aIN

SITU

FORM

TEC

HN

OLO

GIE

S-CL

AIN

SUU

SN

C12

.133

90.

7n/

an/

a12

.96.

44.

536

.19.

86.

0n/

a18

.613

.9n/

aM

UEL

LER

WA

TER

PRO

DU

CTS

INC-

AM

WA

US

NC

6.54

193

0.6

0.6

n/a

3.8

4.5

n/a

5.7

7.0

n/a

15.6

15.5

12.5

2.00

NO

RTH

WES

TPI

PEC

OM

PAN

YN

WPX

US

Buy

30.9

283

1.0

0.8

0.7

8.6

5.9

4.9

9.7

6.4

5.4

13.3

9.9

8.9

n/a

PEN

TAIR

INC

PNR

US

NC

26.6

2,63

51.

11.

0n/

a7.

87.

1n/

a9.

6n/

an/

a12

.511

.610

.20.

98R

OPE

RIN

DU

STRI

ESIN

CR

OP

US

NC

37.4

3,34

91.

91.

8n/

a7.

77.

1n/

a9.

3n/

an/

a13

.211

.710

.30.

87TY

CO

INTE

RN

ATI

ON

AL

LTD

TYC

US

NC

25.5

12,0

840.

80.

80.

65.

54.

93.

89.

77.

85.

9n/

a8.

57.

8n/

aW

ATT

SW

ATE

RTE

CH

NO

LOG

IES-

AW

TSU

SN

C24

.872

50.

60.

60.

65.

05.

74.

96.

66.

96.

112

.413

.111

.7n/

aAverage

0.9

0.9

0.6

7.2

5.6

4.1

10.9

7.0

5.2

13.9

11.3

9.8

0.86

WaterSolutions&Treatment

AR

CH

CH

EMIC

ALS

INC

AR

JUS

NC

24.2

602

0.5

n/a

n/a

4.8

n/a

n/a

7.1

n/a

n/a

16.8

11.1

9.1

0.39

ASH

LAN

DIN

CA

SHU

SB

uy23

.31,

465

0.1

n/a

n/a

n/a

n/a

n/a

n/a

n/a

n/a

6.4

8.1

7.4

n/a

CA

LGO

NC

ARB

ON

CO

RPCC

CU

SN

C15

.262

41.

9n/

an/

a15

.3n/

an/

a25

.3n/

an/

a38

.827

.621

.40.

61D

AN

AH

ERC

ORP

DH

RU

SN

C52

.016

,572

1.8

1.3

1.1

10.1

7.2

5.9

11.6

8.3

6.8

11.8

11.8

10.6

1.77

HER

CU

LES

INC

HPC

US

Hol

d13

.91,

561

1.0

1.0

0.8

6.0

5.5

4.4

8.5

7.2

5.8

8.8

9.2

7.9

0.98

LAY

NE

CH

RIS

TEN

SEN

CO

MPA

NY

LAY

NU

SN

C17

.533

80.

40.

20.

12.

91.

81.

04.

93.

11.

77.

96.

76.

40.

85N

ALC

OH

OLD

ING

CO

NLC

US

Buy

12.4

1,74

21.

31.

11.

06.

96.

55.

610

.19.

27.

513

.710

.68.

40.

37TE

TRA

TEC

HIN

CTT

EKU

SN

C15

.994

50.

60.

70.

59.

57.

05.

411

.08.

26.

219

.915

.913

.70.

80Average

0.9

0.9

0.7

7.9

5.6

4.4

11.2

7.2

5.6

15.5

12.6

10.6

0.83

Automation

EMER

SON

ELEC

TRIC

COEM

RU

SN

C31

.424

,372

1.2

1.1

0.9

6.5

5.8

4.9

7.7

6.9

5.7

11.7

10.2

9.4

1.28

HO

NEY

WEL

LIN

TERN

ATI

ON

AL

INC

HO

NU

SB

uy28

.821

,440

0.8

0.7

0.6

6.0

5.1

4.1

7.3

6.3

5.0

9.0

7.5

6.6

0.58

RO

CKW

ELL

AU

TOM

ATI

ON

INC

ROK

US

NC

25.5

3,71

40.

80.

70.

64.

33.

93.

44.

94.

43.

72.

76.

36.

5n/

aAverage

0.9

0.8

0.7

5.6

4.9

4.1

6.6

5.8

4.8

7.8

8.0

7.5

0.93

WaterMeters

BA

DG

ERM

ETER

INC

BMIU

SN

C29

.743

51.

9n/

an/

a11

.9n/

an/

a14

.5n/

an/

a25

.617

.615

.6n/

aIT

RON

INC

ITR

IUS

NC

67.8

2,32

62.

61.

51.

220

.69.

97.

9n/

a21

.114

.5n/

a19

.615

.6n/

aAverage

2.3

1.5

1.2

16.3

9.9

7.9

14.5

21.1

14.5

25.6

18.6

15.6

n/a

AVERAGE

1.6

1.1

0.8

8.6

5.9

4.6

11.9

8.4

6.4

13.1

12.4

10.8

0.98

Company

Ticker

Price

MktCap

Rating

PEG

EV/Sales

EV/EBITDA

EV/EBIT

P/E

Sou

rce:

Blo

ombe

rg,J

effe

ries

&C

ompa

nyIn

c.es

timat

es

Alt

ern

ativ

e S

olu

tio

ns

- W

ater

Ple

ase

see

impo

rtan

t dis

clos

ure

info

rmat

ion

on p

ages

208

- 2

10 o

f thi

s re

port

.P

age

150

of 2

12


EventJefferies Research has collaborated to provide an overview of thechallenges and opportunities inherent in full-scale deployment ofcarbon sequestration.

Key Points• Cost of carbon capture will be high. Many estimates assume

the capture of carbon emissions will range from $40-70/ton ofCO2 from the current fleet of coal-burning plants. Additionally,the high energy intensity of the capture process could absorbas much as 25% of the output of the underlying plant.

• Technology could be a solution as many companies andgovernments are developing alternative capture processesdesigned to lower cost. Additionally, many are looking atalternatives to coal combustion (coal gasification in IGCC plantsfor example) as another method of reducing carbon emissions.However, many of these potential solutions are either unprovenor too costly for mass deployment at this stage.

• Clarity on carbon costs will be an increasingly importantissue. Emitters will require a clear view on the potential cost of aton of CO2 (either by cap and trade or as a tax) in order tomake informed investment decisions.

• Transportation of CO2 will be an issue as high-pressure CO2cannot be transported in the existing low-pressure natural gasnetwork, even assuming there was capacity to carry both.Shifting captured CO2 from its source to its eventual storageplace will require significant investment in new pipe networksoften covering new routes.

• Storage of CO2 is another hot topic as not only are someproposed storage sites controversial (e.g., deep ocean storage)but we must be certain that, once stored, the CO2 will notsimply seep back to the surface and create yet more problemsfor future generations.

October 6, 2008

Clean Technology Clean TechnologyCarbon Sequestration Primer

Investment SummaryCarbon sequestration is a catch-all term for the capture ofharmful CO2 emissions, principally from coal-fired power andindustrial plants, and subsequent transport for eventual storageor use in industrial or energy applications. Carbon sequestrationthrows out many challenges due to its capital intensity, impact ofpower prices, and transport and storage unknowns.






Carbon Sequestration: An Overview

There is growing interest in carbon sequestration, which is basically a suite of technologies for moving a cloud of CO2 underground. As a solution, this stands in sharp contrast to alternative power technologies that use carbon more efficiently, a bit like going to the gym to work off that second dessert. Carbon sequestration, in our view, is apt to be capital intensive, much delayed, and hostage to the price put on CO2 emissions by governments or capital markets. Indeed, we do not expect the first standardized carbon sequestration systems to be established before 2025 at the earliest, and legal hurdles could push this back significantly. Volatility in the arbitrage between coal and other fuel sources (esp. natural gas and biomass) could also undermine project economics without significant government guarantees on project returns. In the near term, we expect proof-of-concept projects to be lucrative mostly for equipment and technology providers, such as Praxair (PX, Buy, $105 PT).

Other companies that might benefit include the following:

EXHIBIT 1: KEY INVESTMENT THEMES

Value Chain Companies involved

Demonstration Projects

AES, BP, Chevron, ConocoPhilips, EnCana, GE, IMC Global, Norsk Hydro, Rio Tinto, RWE, Shell, Siemens, Spectra Energy, Statoil, Sumitomo Chemicals, Total, TransAlta

Industrial gases Air Liquide, Air Products, Airgas, Linde, Praxair Separation technology Upstream chemicals, process catalysts (BASF, JohnsonMatthey, Umicore), Alstom, Honeywell

Gasification technology

Air Liquide, ConocoPhilips, Eastman Chemical, GE, Linde, Mistubishi Praxair, Siemens, Royal Dutch Shell

E&C Fluor, Shaw Group, Washington Group Derivative Themes Valves & pumps, stainless steel, natural gas firms, utilities, oilfield service companies


EXHIBIT 2: DOE CARBON SEQUESTRATION ROADMAP: TESTING OPTIONS THROUGH 2022

1997 Carbon Sequestration Program started

2001 First Annual Carbon Sequestration National Conference

2003 Regional Carbon Sequestration Partnerships Program started

2005 Started Validation Phase of Regional Carbon Sequestration Partnerships

2006 Characterization Phase completed

2007 Identify capture technologies that increase cost of energy services less than 20% for pre-combustion systems and 45% for post-combustion systems and oxy-combustion systems

Start Development Phase

2008 Develop protocols that enable 95% of stored CO2 to be credited as net emissions reduction

2009 Complete Validation Phase

2011 Initiate at least one large scale (>1m tpy CO2) storage demonstration in a geologic formation

2012 Develop fossil fuel conversion systems that capture 90% of CO2 with 99% storage permanence at less than 10% increase in cost of energy services

2014 Develop protocols that enable 99% of stored CO2 to be credited as net emissions reduction At least two slipstream tests of novel CO2 capture technologies that offer significant cost reductions

2015 Develop terrestrial sequestration technologies to lower cost to $5/t of carbon or less

2016 Begin one saline sequestration project with brine water recovery

2018 Begin large-scale testing of novel CO2 capture technologies

2020 Field demonstration of at least one technology for enhancing CO2 mineralization in-situ

2022 At least one large-scale demonstration of integrating biomass gasification and coal gasification to sequester carbon

Source: DOE

Clean Technology


Page 152 of 212

How it works

Carbon sequestration involves a five-step process, not all of which has been demonstrated at scale yet.

1. Capture the carbon.

The most intuitive approach is to filter out CO2 from smokestacks, like a cigarette filter snaring tar and fine particles. Various solvent and membrane technologies are being developed to improve the efficiency of this process, but almost all are energy intensive. Flue gas in a smokestack might be 10%–12% at a coal-fired power plant or 3%–20% at an industrial facility. This has the significant advantage of being able to be rolled out quickly on existing capacity. Key technologies in this phase include:

• Chemical solvents, such as amines (the only commercial solution), advanced amines (pilot scale), and aqueous ammonia (laboratory scale).

• Physical solvents such as ionic liquids, and sorbents such as metal organic frameworks (both at laboratory scale). • N2/CO2 membranes, based on membrane/amine hybrid technology or enzymatic CO2 processes (laboratory

scale).

A major challenge for carbon capture is the high power intensity and cost of the capture process. A typical amines-based solution has a vampiric consumption of as much as 20%–25% of the total power produced at the plant and will reduce the external energy efficiency of a typical coal fired generator from 35%–43% to 25%–33%. This lower output, combined with higher capital and operating expenditure, also will increase the production cost of coal-fired energy. If we assume fully loaded carbon capture cost of $40/ton, the cost of coal-fired electricity will increase from $0.045– $0.05/kWh to $0.07-$0.08/kWh. These costs cover only the capture of the carbon as transportation and storage will vary depending on site location.

Another approach is to inject oxygen into the boiler to burn the coal more efficiently. The flue gas at oxy-fuel coal-fired boilers can be as much as 90% CO2. The downside, however, is that the plant requires an adjacent oxygen plant and can be 10%–15% less energy efficient due to the need to separate out the CO2. As a rule of thumb, each 400 MWh oxyfuel facility would require 500 tpd of oxygen. The current approach involves enhancing conventional oxyfuel processes with recirculation, but companies are also working on next-generation processes involving oxygen-transport membranes and chemical looping. Currently, four IGCC power plants have been established worldwide, although these are demonstration projects dependent on government support due to their high capital and operating costs.

EXHIBIT 3: GLOBAL SYNGAS LANDSCAPE, 2007

Sources Co-Products H/CO Demand Capacity Coal 55% Slag Chemicals, Fertilizer 45% 56,238 MW

Heavy Oil 33% Mercury Fisher-Troppes Liquids 28% South Africa 27% Natural Gas 8% Sulfur Electric Power 19% China 24%

Petcoke 2% Liquid Ar, N2, O2 Gaseous Fuel 8% Europe 24% Biomass 2% North America 14%

Source: NTEL, DOE, Madison Power, Jefferies & Company, Inc. estimates

The third, most complex phase is to try to extract the CO2 at the start. Coal or other materials are turned into syngas (H2 + CO), then water (H2O) is added to make a mix of hydrogen and CO2. The CO2 stream is separated out, and then the H2 combusted. In IGCC plants, as much as 80% of the CO2 can be captured without a significant yield loss. Pre-combustion technologies include the following:

• Chemical solvents, mostly amines. • Physical solvents, including glycol and methanol. Ionic liquids are an area of active research. • Chemical and physical sorbents based on metal organic frameworks. • Next-generation membranes based on polymeric, ceramic, and hollow fiber materials. • CO2 hydrates are also being explored at the laboratory scale for gas separation.

In all three cases, other gases, such as nitrogen and oxygen, are separated out and either released to the atmosphere or diverted as industrial co-products. The CO2 is compressed and purified to extract water and gaseous impurities. In general, these technologies are expected to remove 80%–95% of the CO2 emitted by power plants or industrial facilities.

Clean Technology



Page 153 of 212

EXHIBIT 4: INDUSTRY PARTICIPANTS IN DOE CARBON CAPTURE ROADMAP

Pre-combustion Oxyfuel Post-combustion Nexant: CO2 hydrate process for gas separation

Praxair: Advanced oxy-fuel boilers and process heaters

RTI: Dry regenerable CO2 sorbents

NETL: Sorbents for CO2 separation and removal Alstom Power: Oxygen firing in coal-fired boiler NETL: CO2 scrubbing with

regenerable sorbents MPTC: Selective ceramic membranes

Foster-Wheeler: Oxygen-enriched pet coke-fired boiling system with CO2 sequestration

NETL: Novel amine-enriched sorbents

BP: CO2 capture project NETL: Modular CO2 capture facility

Eltron: Membranes UT AustinL Adsorption with K2CO3/Piperazine

Air Products: Enhanced sorption, water-gas shift technology

INEEL, LANL: High-temperature polymer membranes FuelCell Energy: Direct fuel cell

Source: DOE, Jefferies & Company, Inc. estimates

2. Transport The liquid CO2 can be shipped by pipeline or truck. Currently approximately 3,600 miles of CO2 pipelines transport roughly 42m tpy of CO2 (2bn scfd), mostly for enhanced oil recovery applications (EOR). EOR uses CO2 to flush out oil and natural gas from deposits. This can increase extraction efficiency by 35%–60%. Some CO2, however, is left behind in the pores of the rock. This was once viewed as a bad thing: expensive CO2 left behind. Now it is starting to be viewed more positively, although the economics remain hazy. Moreover, the scale could be daunting. While some studies have suggested that more than 75% of the CO2 captured could be stored in reservoirs directly below the emissions, other studies have estimated that a carbon sequestration platform might need a 10x–50x expansion of these high-pressure pipes (and lower-pressure pipes already in place for natural gas cannot be switched over). This could imply a significant investment in pipes, of course — but more importantly could run into bottlenecks such as valves, pumps, and key metal supplies. There could also be significant legal hurdles to establishing appropriate rights of way. For example, there is already a debate as to whether CO2 should be classified as a commodity (the market for it already exists) or as a pollutant (which would pit the EPA against the BLM and FERC).

Another issue is that, even if there were capacity high pressure CO2 cannot be transported in the existing low pressure natural gas pipeline network. Brand new high pressure pipe networks would have to be constructed, often times on brand new routes as the current infrastructure is designed to bring natural gas from its source to the power plant while the CO2 network may not return the product to the original location. A simple way to think of this is that the carbon, in the form of coal, is transported by rail from the mine to the power plant while the gas would likely return (for example, assuming coal seam storage) by a parallel pipeline.

Liquid CO2 can also be distributed by truck, in what is called the merchant or bulk liquid market. This now represents a roughly 13.1m tpy (36,000 tpd) business. Roughly two-thirds of industrial CO2 is used in the food & beverage industry for making, processing, and packaging products. While sales to the soft drink market have flattened out, niche markets such as environmentally friendly cleaning solutions and water treatment have picked up the slack due to tighter environmental standards. Taken together, these markets lead to a roughly 2.0%–2.5% growth rate. The main players in the merchant market are Linde (31%), Airgas (ARG, Buy, $75 PT; 17%), Praxair (15%), Air Liquide (11%), and EPCO (8%).

Clean Technology



Page 154 of 212

EXHIBIT 5: MERCHANT CO2 SUPPLY, 2006

Ammonia22%

Other3%

Wells24%

Refineries26%

Ethanol25%

Source: Cryogas International

3. Storage

So far, the steps we have discussed involve well-established technologies and processes. The new step is toattempt to store the coal in a permanent fashion — basically moving a cloud underground. Besides oil and gasreservoirs, one suggestion is to use coal seams — if they are economically unattractive to dig out, they might aswell be used as sponges to absorb CO2 from other coal plants. Another scenario is to pump the CO2 into the deepocean, on the theory that the pressure in the depths (below 10,000 feet) will keep the CO2 trapped in liquid form.

Who is involvedThe appeal of a carbon sequestration project, in our view, remains contingent on the quality of the carbon source(volumes, purity, pressure, reliability), the distance to the storage area, the cost of electricity, and the reliability ofthe storage area and related monitoring costs. To make the economics work, the focus is on the highest-volumepoint sources of CO2: coal plants. We estimate that approximately 90% of global industrial CO2 emissions comefrom industries that are also intensive customers for the industrial gas companies.

EXHIBIT 6: GLOBAL INDUSTRIAL CO2 EMISSIONS (%), 12.76BN TPY TOTAL

PowerGeneration

61%

Other Industry10%

Oil Refining5%

CementManufacturing

9%

Iron and Steel11%

Petrochemicals4%

Source: IEA, Praxair

Not surprisingly, then, the first focus for investors has been on companies involved in power generationdemonstration projects and their industrial gas suppliers. In the first group, one finds companies such as AmericanElectric Power (AEP, Buy, $42 PT), BP, E.ON, General Electric (GE, Hold, $30 PT), RWE, Siemens, Shell, Statoil,as well as their E&C partners. In the second, one finds Praxair and Air Liquide. These companies have announceddemonstration facilities that, by 2016, aim to generate almost 10,000 MW of power using various carbon captureand sequestration technologies.

Clean Technology



Page 155 of 212

EXHIBIT 7: EXPECTED CARBON CAPTURE SEQUESTRATION DEMONSTRATION PROJECTS (CUMULATIVE MW*)

1,0002,0003,0004,0005,0006,0007,0008,0009,000

10,000

2009 2010 2011 2012 2013 2014 2015 20160

1,0002,0003,0004,0005,0006,0007,0008,0009,000

10,000

2009 2010 2011 2012 2013 2014 2015 20160

Source: IEA, Jefferies & Company, Inc. estimates * 500 MW = 1.4m tpy of coal = 3.5m tpy of CO2

According to DOE forecasts, CO2 capture and sequestration technologies are expected to contribute almost 40% of the targeted reduction in GHG emissions by 2050, or a net 15% reduction in industrial CO2 emissions. Such targets are only likely to increase as forecasts are rolled forward to accommodate an expanding economy.

Location

The first factor to consider is storage capacity. A survey of 41 states and four Canadian provinces identified between 1.1–3.6 trillion tpy of CO2 storage capacity in different geologic formations. The same survey identified 4,365 stationary sources of carbon that produce an estimate 3.81bn tpy of CO2, or roughly 48% of total U.S. CO2 emissions. In other words, the survey suggested potential storage capacity for 304–956 years of CO2 emissions at 2005–2006 emission rates.

EXHIBIT 8: GEOLOGIC SEQUESTRATION CAPACITY (BN TPY OF CO2)

Storage Capacity U.S. Low U.S. High Global Deep Saline 919 3,378 1,000-10,000 Oil & Gas Reservoirs 82 82 675-900 Unminable Coal Seams 156 184 5-200 Total 1,158 3,644 1,680-11,100 Multiple of current emissions (incl. Non-stationary) 148x 467x 56x-350x

Source: DOE, IPCC

• Deep saline storage: The most significant opportunity for carbon sequestration involves pumping CO2 into layers of porous rock that are saturated with brine. While initial estimates for storage capacity are large (0.9–3.4bn tpy), the survey made some broad assumptions, particularly that: 1) any depth of more than 2,500 feet would provide sufficient pressure and temperature to keep the CO2 in a dense (supercritical) or liquid state; and 2) CO2 would occupy 20%–80% of the area inventories and 25%–75% of the estimated available thickness of the rock formation.

EXHIBIT 9: OPTIMAL CO2 SEQUESTRATION UNDERGROUND (CO2 KG/CUBIC METER)

20

25

30

35

40

45

50

0

2500

5000

7500

1000

0

1250

0

1500

0

Depth (ft)

CO

2bu

lkvo

l.re

sidu

al(k

gm/m

3)

Maximum storage capacity

20

25

30

35

40

45

50

0

2500

5000

7500

1000

0

1250

0

1500

0

Depth (ft)

CO

2bu

lkvo

l.re

sidu

al(k

gm/m

3)

Maximum storage capacity

Source: DOE, Jefferies & Company, Inc. estimates

Clean Technology



Page 156 of 212

• Unminable coal seams: Some coal seams are too deep or too thin to be worth mining. Injecting CO2 into the coal provides an effective storage medium, with the benefit that for every 3–13 molecules of CO2 injected, a molecule of methane can be released, depending on the seam. Organic rich shales, which contain 1%–2% organic materials, could also provide an attractive carbon sink. In some regions, most notably Alberta, these shales are actually being extracted rather than used as storage capacity. In order for this solution to benefit the environment, the target coal seam must have methane collection infrastructure for use and or flaring, else the released methane molecule would damage the ozone layer.

• Oil & gas reservoirs. CO2 can be injected into oil and gas reservoirs, and then be trapped below a layer of impermeable rock. If a typical oil project recovers 30%–40% of the original oil in place, injecting CO2 can free up another 10%–15%. This is estimated to be viable around $30–$40/bbl crude oil. One challenge, of course, is that one wants the CO2 to stay in the ground, rather than come back up with the oil.

• Deep ocean: There are several proposed technologies for storing CO2 in the depths of the ocean, where pressure would theoretically keep the CO2 trapped in a liquid state. The prospect of deliberately acidifying the ocean, however, could well provoke resistance from environmentalists concerned over the viability of the food chain. Estimates for the total capacity to store CO2 in the oceans range from 1,400–20,000,000bn t.

EXHIBIT 10: CO2 ABSORPTION BY FORESTRY (M TPY AS OF 2030) DEPENDING ON CARBON PRICES ($/T OF CO2)

m tpy U.S. Europe Total $20/t

Afforestation 133.5 35.7 1,618.0 Reduced Deforestation 2.0 1.7 2,133.0 Forest Management 413.4 51.0 1,965.2 Total 548.9 88.4 5,716.2

$50/t Afforestation 267.0 63.3 2,750.6 Reduced Deforestation 5.0 4.4 3,239.0 Forest Management 922.2 83.3 3,583.6 Total 1,194.2 151.0 9,573.2

$100/t Afforestation 445.0 115.0 4,045.0 Reduced Deforestation 10.0 10.0 3,950.0 Forest Management 1,590.0 170.0 5,780.0 Total 2,045.0 295.0 13,775.0

Source: IPCC * Note: $1/MSCF = $17/t of CO2

• Surface: To supplement the potential geologic storage capacity, U.S. agencies have also evaluated surface sequestration through improved agricultural practices (potentially 18–20m tpy), mineland reclamation (2m tpy), wetlands (1.5m tpy), and silvaculture (85m tpy). Similarly, the IPCC (Intergovernmental Panel on Climate Change) estimates that afforestation could capture roughly 133m tpy of CO2 by 2030, at $20/t of CO2 (and 3x that at $100/t). Globally, the IPCC estimates that afforestation could capture 1.6bn tpy of CO2 at $20/t CO2. This compared with reducing emissions by 2.1bn tpy by reducing the pace of deforestation, and another 2.0bn tpy from improved forest management techniques.

Clean Technology



Page 157 of 212

EXHIBIT 11: KEY ISSUES FOR DIFFERENT CARBON STORAGE OPTIONS

Property Terrestrial biosphere Deep ocean Geological reservoirs CO2 sequestered or stored

Stock changes can be monitored over time

Injected carbon can be measured Injected carbon can be measured

Ownership Stocks will have a discrete

location and can be associated with an identifiable owner

Stocks will be mobile and may reside in international waters

Stocks may reside in reservoirs that cross national or property boundaries

and differ from surface boundaries

Management decisions Storage will be subject to

continuing decisions about land-use priorities

Once injected there are no further human decisions about

maintenance once injection has taken place

Once injection has taken place, human decisions about continued storage

involve minimal maintenance, unless storage interferes with resource

recovery

Monitoring Changes in stocks can be monitored

Changes in stocks will be modeled

Release of CO2 can be detected by physical monitoring

Expected retention time Decades, depending on management decisions

Centuries, depending on depth and location of injection

Essentially permanent, barring physical disruption of the reservoir

Physical leakage Losses might occur due to

disturbance, climate change, or land-use decisions

Losses will assuredly occur as an eventual consequence of

marine circulation and equilibration with the

atmosphere

Losses are unlikely except in the case of disruption of the reservoir or the

existence of initially undetected leakage pathways

Liability A discrete land-owner can be

identified with the stock of sequestered carbon

Multiple parties may contribute to the same stock of stored

CO2 and the CO2 may reside in international waters

Multiple parties may contribute to the same stock of stored CO2 that may lie

under multiple countries

Source: DOE, IPCC, Jefferies & Company, Inc. estimates

Cost

The second factor constraining policy choices is the incremental cost of capturing and storing the CO2. The front end of the CO2 cost chain is well established. First, capture CO2, which is typically 3%–20% of waste gas at industrial facilities and as much as 90% of flue gas at oxy-fuel coal fired boilers. Separate out the nitrogen and oxygen, which is either released to the atmosphere or diverted for industrial uses. Then compress and purify the gas to extract water and gaseous impurities. Then ship the CO2 via truck or pipeline. The new step for the industry is to attempt to store the coal in a permanent fashion—basically moving a cloud underground. To make the economics work, the focus is on the highest-volume point sources of CO2 that emit the most concentrated CO2 streams: coal plants, oxy-fuel boilers, and IGCC plants.

On a cost basis, most estimates peg CO2 capture at $40–$70/t, with additional costs accrued for compression, purification, and storage. Some next-generation technologies are expected to lower the cost of CO2 capture to $5–$25/t. The IPCC cites estimates that the cost of capture and compression could be reduced 20%–30% over the next few decades through innovation.

EXHIBIT 12: RELATIVE APPEAL OF DIFFERENT SOURCES OF CO2

SMR22Cement

SMR1NGCC

EthyleneRefinery Steel

PC

IGCC

Oxy-fuel PC Boiler

Coal to SNGAmmonia

SMR3

Ethanol

0102030405060708090

100

0 2000 4000 6000 8000 10000

CO2, Mt/day

CO

2C

once

ntra

tion

%(V

/V)

Too Small

Too Dilute

Cost-effective

SMR22Cement

SMR1NGCC

EthyleneRefinery Steel

PC

IGCC

Oxy-fuel PC Boiler

Coal to SNGAmmonia

SMR3

Ethanol

0102030405060708090

100

0 2000 4000 6000 8000 10000

CO2, Mt/day

CO

2C

once

ntra

tion

%(V

/V)

Too Small

Too Dilute

Cost-effective


Clean Technology



Page 158 of 212

EXHIBIT 13: CO2 CAPTURE COSTS (2004-2005 COST BASE)

Source % CO2 Capture (tpd) Capital ($m) CO2 Cost ($/t) Capture

Hydrogen plant Syngas 20% 1,400 $35$40

$5-$55 (IPCC)

IGCC Syngas 40% 6,500 $100 $30

$15-$75 (IPCC) Coal fired boiler 13% 1,100 $42 $55Oxy-fuel coal fired boiler 90% 6,500 $200 $35 Gas turbine 3.50% 1,100 $60 $70Capture at cement plant 20% $25-$45 Capture from other industrial sources 3%-20% $25-$115

Compression & Purification $10-$20

Transportation (250 km by pipe) Pipe (250km) $1-$8 Off-shore pipe (250km) $2-$12

Up front capital cost ($/mile of pipe) (labor 26%, materials 25%, rights of way 14%) $1-$2m/mile Shipping $2-$25

Storage Geologic storage & monitoring $0.60-$1.30 Mineral carbonation $50-$100 Ocean Storage (no monitoring, pipe)* $5-$30 Ocean Storage (ship/platform)* $12-$15

Total Cost $32-$145 Source: Jefferies & Company, Inc. estimates, IPCC, Congressional Research Service *Assume 100-300 miles offshore

EXHIBIT 14: KEY ASSUMPTIONS FOR ALL-IN CO2 COST ESTIMATES BY EMISSION SOURCE

$/Mt CO2 Separation Compression Pipeline Uncertain Total Key Assumptions: Ammonia $0.00 $15.20 $7.20 $4.00 $26.40 SMR/Amine $0.00 $16.80 $11.20 $4.00 $32.00

- Capture from flue gas by amine

Coal Power $32.40 $11.20 $3.20 $8.00 $54.80 - Capital recovery: 14%/yr Ethanol $0.00 $22.40 $30.40 $5.20 $58.00 SMR/PSA Syngas $26.80 $15.20 $11.20 $5.20 $58.40

- Additional $2/Mt for transport in major pipeline

SMR/PSA Fluegas $41.20 $14.40 $7.20 $5.60 $68.40 Ethylene $48.00 $13.60 $4.80 $4.80 $71.20

- 25 mile pipeline to major CO2 pipeline

NG Power $50.80 $12.40 $5.20 $4.80 $73.20 Cement $43.60 $15.20 $10.00 $7.20 $76.00

- COE: $0.05/kWh for utilities and $0.06/kWh for others

Integrated Steel $50.40 $12.80 $3.20 $10.40 $76.80 - Natural Gas: $6/MMBtu HHV Refinery $54.40 $13.60 $4.00 $8.00 $80.00 - Coal: $1.5/MMBtu HHV

Source: Praxair, Jefferies & Company, Inc. estimates

The costs that are typically excluded from analysis, such as pipeline costs, are not negligible. To sequester 50% of the stationary emissions from the U.S., for example, and assuming no need to build long-distance pipelines or resort to ocean storage, with an average cost of $55/t, would cost around $100bn a year, or 0.8% of GDP. To build out a pipeline 50x the current network in the U.S. would cost another $180–$360bn in capital outlays, assuming no impact on the price of steel, molybdenum, valves, pipes, and other materials. Similarly, property rights and related litigation would surely increase this outlay.

Clean Technology



Page 159 of 212

EXHIBIT 15: KEY ASSUMPTIONS FOR ALL-IN CO2 COST ESTIMATES FOR NEXT-GENERATION TECHNOLOGIES

$/Mt CO2 Separation Compression Pipeline Uncertain Total Coal to syngas/H2 $0.00 $13.10 $5.79 $10.34 $29.24 Oxyfuel $24.83 $13.10 $3.45 $4.83 $46.21 IGCC $19.31 $7.59 $3.10 $10.34 $21.03


EXHIBIT 16: RULE-OF-THUMB CO2 EMISSIONS BY SOURCE (TPD)

Source Capacity Mt CO2/day Coal Power 500 MW 983

Integrated Steel 2m tpy 933Refinery 200,000 bpd 522 Ethylene 1,600m lb/yr 467

Natural Gas Power 500 MW 389 Hydrogen ASU 100m scfd 222

Ammonia 2,000 tpd 222 Cement 2,000 tpd 144Ethanol 50m gal/year 39


EXHIBIT 17: RELATIVE APPEAL OF DIFFERENT COGENERATION PLANT TECHNOLOGIES

Technology Fuel Capacity MW Electrical efficiency (%) Overall efficiency (%) Steam turbine Any combustible 0.5-500 17-35 60-80 Gas turbine Gaseous & liquid 0.25-50+ 25-42 65-87

Combined cycle Gaseous & liquid 3-300 35-55 73-90 Diesel and Otto engines Gaseous & liquid 0.003-20 25-45 65-92

Mirco-turbines Gaseous & liquid 0.05-0.5 15-30 60-85 Fuel cells Gaseous & liquid 0.003-3+ 37-50 85-90

Sitrling engines Gaseous & liquid 0.003-1.5 30-40 65-85

Source: IPCC

Longer-term risks

Cost recovery. Given the significant capital costs that could be involved with installing a new carbon pipeline system, cost recovery could emerge as a significant issue. For example, would utilities pay for the cost through raising electricity rates? Would carbon credit markets be willing to fund a project years before any liquid CO2 flows down the pipe? If intrastate pipelines are later connected to an interstate pipeline, how would access, terms of service, and fee structures be handled?

Ownership rights. Does carbon sequestration capacity go to the owner of surface rights or mineral rights? If an accident in a saline aquifer storage scheme increases the acidity of water nearby, to whom would the owners of water rights have recourse?

Tax incentives: Will there be a tax credit for pipeline construction similar to the 15% income tax credit for approved EOR technologies? In the 1970s, EOR projects were exempt from oil price controls: will a similar incentive be needed this time for lenders to support the projects?

Flexibility: Will pipeline owners be able to arbitrage between the industrial value of CO2 and the carbon credit value of sequestration, which may defeat the purpose of a sequestration project—and if not, how will compliance be monitored?

Scale. To date, all of the carbon sequestration projects have been demonstration projects. There could be significant complications when investors attempt to scale up these projects.

Monitoring. One risk is that the wells used for monitoring CO2 levels, to verify that the storage process has been effective, are also the weak points in the storage system. These wells could corrode over time, or prove more vulnerable to seismic shocks.

Accidental release. From 1986 through 2006, there were 12 leaks reported from CO2 pipelines, though no fatalities. If CO2 pipelines achieve a scale comparable to natural gas pipelines, will the accident rate converge on

Clean Technology



Page 160 of 212

that of natural gas pipelines (roughly 550 incidents a year)? Another longer-term risk is that eagerness to sequester CO2 leads to projects that store the CO2 in inadequate areas. What might make an area inadequate? Lack of a stable cap, for example, as one might find in granite domes, or vulnerability to movements in the Earth’s crust. In either case, will the pipelines be indemnified from criminal or civil liability? Under one scenario, the CCS industry could follow the path of the nuclear industry, with utilities indemnified of the damages arising from any significant accidental release.

Reliance on innovation. Beside the economic and regulatory uncertainties, many of the technologies involved in carbon sequestration remain in early stages of development. This could lead to further delays, and complicate the financing of, carbon sequestration projects.

EXHIBIT 18: MATURITY OF DIFFERENT COMPONENTS IN THE PROPOSED CARBON SEQUESTRATION SYSTEMS

Component Technology Limited

Mature Capture Post-combustion x

Pre-combustion xOxyfuel combustion x

Industrial separation (natural gas, ammonia) xTransportation Pipeline x

Shipping xGeological storage Enhanced oil recovery (EOR) x

Gas or oil fields xSaline formations x

Enhanced coal bed methane recovery xOcean storage Direct injection/dissolution x

Direct injection/lake xMineral carbonation Natural silicate minerals x

Waste minerals xIndustrial uses for CO2 x

Source: IPCC

Clean Technology

Research Demonstration


Feasibility


Page 161 of 212





EventThis report is designed to assist investors in surveying theadvanced battery technology landscape — reviewing individualtechnologies such as NiCd, NiMH, Li-ion, ultracapacitors andalternative technologies.

Key Points• Rechargeable battery industry expected to grow faster than

global GDP. Industry growth is expected to be driven byincreased demand from developing economies, proliferation ofconsumer devices, and new uses for industrial batteries,including hybrid vehicles and clean energy storage.

• Market share growth for high energy density products. Weexpect that in general, higher energy density products will growfaster than low energy density products. As such, we believethat lead-acid batteries and NiCd are likely to experience slowergrowth as customers switch to higher density products. Thatswitch will, however, be gradual in our opinion given therelatively lower cost of low energy density batteries that willensure their continued use until newer, higher densitytechnologies can offer improved performance and competitivelifetime cost benefit.

• Hybrid vehicle demand a key catalyst of industrial batterygrowth. We expect the hybrid vehicle market to continue togrow rapidly, hitting 5% market share in the U.S. early in thenext decade. While Ni-MH technology is the current standardhybrid solution and likely beneficiary of near term revenuegrowth, Li-ion is the heir apparent technology due to its lowerweight and higher energy density. Lead acid is the obviousloser as traditional batteries lose share.

• Ultracapacitors act as a supplement to batteries. We believethat ultracapacitors are likely to grow faster than the industry asthese technologies are just now being designed into products.In our opinion, the fast charge/discharge capabilities forultracapacitors make them likely candidates to be paired withstorage capacity of batteries in hybrid and wind applications.

October 1, 2008

Clean TechnologyEnergy Storage - Battery Technology

Clean TechnologyAdvanced Battery Technology Primer

Investment SummaryWe expect the advanced battery technology market to growfaster than GDP due to strong growth in consumer electronics,wireless telecom and hybrid vehicles. Along with companies withexposure to those markets, we would recommend investing incompanies that have exposure to higher energy density products.

Alex Barnett, CFA, Equity Analyst00 33 1 5343 6714, [email protected]


A Primer on Batteries

Given the trend towards increased power usage in electronic and industrial applications and potential in hybrid vehicles, we believe it is appropriate to provide greater detail on the various battery technologies and their growth prospects. This primer focuses on rechargeable batteries, which account for only 10% of unit volume, but importantly, 63% of industry revenues. The rechargeable market is segmented into three main areas, Transportation, Industrial, and Portable applications. While we believe the sector as a whole will grow in line, or slightly faster than GDP, within each segment, we expect a shift in battery usage from lower density batteries to higher density, which is likely to result in faster than industry growth for those companies with advanced battery technologies.

GLOBAL BATTERY MARKET

$50b+ Global Battery Market

Non-Rechargeable

37%

Industrial17%

Portable16%

Transport30%

Rechargeable Batteries by Type

0%

15%

30%

45%

60%

75%

Lead-Acid Li-ion Ni-MH Ni-Cd Other

%of

Tota

lSal

esVa

lue

Source: Laboratoire de Reactivite et de Chimie des Solides, Freedonia, www.batteryuniversity.com

Batteries – A Quick Overview To start from the basics, a battery is a stand-alone power source that converts chemical energy into electrical energy through chemical reactions. A battery consists of a positive and negative electrode and an electrolyte. Energy is created when chemical reactions transfer energy from the electrodes to the electrolyte at their interface.

There are two key types of batteries — primary and secondary. Chemical reactions in primary batteries are irreversible, which makes them non-rechargeable. Primary batteries account for roughly 90% of total global battery unit volume, mainly small consumer disposable batteries. In secondary batteries, the chemical reaction can be reversed, which allows for recharging the battery, with the number of recharging cycles dependent on the battery’s material. Secondary batteries account for roughly 10% of global battery production by volume.

A battery’s characteristics depend, in large part, on their constituent materials. There are a wide range of battery systems made from varied metals, each with distinct attributes. Table 1 illustrates the most common battery systems.

Energy Storage - Battery Technology

Alex Barnett, CFA, Equity Analyst, [email protected], 00 33 1 5343 6714

EXHIBIT 1:


Page 164 of 212

BATTERY CHARACTERISTICS

Lead Acid Re-useableAlkaline NiCd NiMH Li-ion

Gravimetric Energy Density (Wh/kg) 35-50 80 45-80 60-120 110-160Cycle Life (# of cycles) 200-300 50 1500 300-500 500-1000Self-discharge / Month 5% 0.3% 20% 30% 10%Shelf Life (months) 18 6 6 12 18Operating Temperature (Celsius) -20 to 60 0 to 65 -40 to 60 -20 to 60 -20 to 60Battery Cost per Cycle $0.10 $0.10 - $0.50 $0.04 $0.12 $0.14

Key Commercial Uses Industrial/Automotive Electronics Electronics/

IndustrialElectronics/ Automotive

Electronics/ Specialist

Date of Commercial Introduction 1970 1992 1950 1990 1991Toxicity High Medium High Low Low

Source: Battery University, Jefferies Intl. Ltd.

Lead Acid Is the Standard

As Chart 2 indicates, lead acid batteries continue to account for the majority of sales of rechargeable batteries. The lead-based battery technology was first discovered in 1859 and today’s flooded batteries are not significantly different than those made 100 years ago. Lead acid batteries are widely used in automobiles, aircraft, UPS applications, motive power (forklifts), telecom, and reserve power. The benefits of the lead acid technology are that it is proven, with over 100 years of commercial/residential use and it is safe. Additionally, the lead-acid batteries are the cheapest battery technology available. The disadvantage of this battery type is that it has lower energy density in comparison to the newer advanced battery technologies, is significantly heavier, and has greater life cycle constraints.

Outlook — GDP or slightly below: We expect lead acid batteries to continue to grow, but only in line with GDP, or even slightly below it. As Chart 2 indicates, the automotive market is the predominant end market, accounting for two-thirds to three-quarters of all lead-based battery requirements, with the remainder servicing the industrial market. Key battery suppliers to the automotive market include Johnson Controls, Exide Technologies, Matsushita, and Japan Storage Battery. In the industrial market, EnerSys (ENS, $19.71, Buy), Exide Technologies, East Penn, and C&D Technologies are the main players.

LEAD-ACID MARKETS

Lead-Acid Markets

Reserve15%

Automotive73%

Motive12%

Industrial27%

Source: Laboratoire de Reactivite et de Chimie des Solides

Auto Markets to Benefit from China/India Growth, but Hybrids Are a Risk. The demand for automobile batteries is relatively stable in developed markets with roughly 80% of battery purchases coming from replacement batteries. Volume growth is driven by the 20% from new car sales and also from a growing installed base. Global manufacturers of lead acid batteries may benefit from higher growth in China and India as those countries continue to rapidly purchase/build new vehicles; however, given the low barriers to entry in the automotive market, we



EXHIBIT 2:

EXHIBIT 3:


Page 165 of 212

believe that local manufacturers are likely to benefit from this growth at the expense of the larger multinational battery suppliers. Additionally, increased hybrid sales could negatively impact lead acid battery sales. Hybrid vehicles typically use NiMH (or Li-ion) batteries instead of lead acid. Therefore, as hybrid vehicles grow, the resultant demand for lead acid batteries would shrink. We are forecasting ~5% of the U.S. auto market to be hybrid vehicles by 2011, which would subsequently impact demand for lead acid batteries.

Industrial End Markets Expected to Grow Modestly. The industrial market is composed of the motive (forklift) market and the reserve power market. We believe the telecom market is likely to grow in excess of GDP as telecom carriers have started to ramp capex spending to meet the increasing demand for wireless and data services. In motive power, we expect the segment to move in line with GDP in the western nations, while at a faster rate in developing regions like China and India.

NICD MARKET PENETRATION IN INDUSTRIAL APPLICATIONS

Penetration of NiCd Batteries in Industrial Applications

0%

20%

40%

60%

80%

100%

Aviation Rail IndustrialBack-up

TelecomNetworks

Source: Saft

Industrial Markets Relatively Safe, but Alternative Technologies Encroaching. Lead acid batteries are the low-cost battery solution to provide non-CO2 emitting power for forklifts (motive power) and reserve power (i.e., telecom base stations). Because of the higher costs, advanced battery solutions like NiCd (3x more expensive), NiMH (4x more) and Li-Ion (5x+) are less likely to be chosen as replacements for lead acid batteries in the industrial markets. For example, for forklifts, lead acid batteries provide the least expensive working solution and the additional weight of a lead battery, which is a negative in many applications, is here a benefit, acting as a counterweight for whatever the forklift is lifting. In the reserve power market, using lead acid batteries for telecom base stations and in uninterruptible power supplies (UPS) is the lower-cost option. The exception to this rule is in remote (and hot) areas where the shorter life cycle and maintenance demands of lead make costs significantly higher and can make the use of advanced batteries the cheapest initial solution. Saft’s (SAFT FP, €27.96, Buy) recent contract win to replace AT&T’s base stations with NiCd is testament to the slow encroachment of more advanced technology which we believe will be a long-term, but gradual trend.



EXHIBIT 4:


Page 166 of 212

Nickel-Cadmium (NiCd)

NiCd batteries make up a small niche of the industrial battery market with only 4% of total units. NiCd batteries are largely used in industrial applications including industrial and telecom standby power, and in the aviation and rail markets for back-up power and starting systems. Additionally, NiCd batteries are used in consumer electronics and power tools. The batteries have better energy density than lead acid batteries and a longer cycle life. The major disadvantages with Ni-Cd are that it is environmentally unfriendly (Cadmium, like lead, is a recognized carcinogen), and is 1.5x to 3.0x the price of lead acid batteries in industrial applications. Given the higher energy density, improving cost competitiveness and better safety profiles of NiMH and Li-ion, we remain cautious on NiCd. While slowed by current cost considerations we are already seeing conversion to newer technologies — most notably in power tool batteries and consumer electronics for the time being. Key manufacturers of NiCd batteries include SAFT, Hoppecke, and Panasonic for industrial uses and Sanyo, BYD and Saft in the consumer market.

Nickel-Metal-Hydride (NiMH)

NiMH batteries are used in the transportation (hybrid vehicles) and the consumer electronic markets. The advantages of this battery type are that it has higher energy density compared to lead acid and Ni-Cd batteries and is non-toxic. However, it has reduced cycle life compared to Ni-Cd. Importantly, battery planners can design recharging programs such that the battery never fully discharges, thereby extending the overall battery life to an acceptable level. For example, in hybrid battery pack, the engine will kick in when the battery charge has dropped to a prescribed level (~60% charge) and not allow the battery to fully discharge. By combining the NiMH battery with an advanced battery monitoring system, manufacturers have increased the expected battery life to more than 10 years.

Outlook. We believe that NiMH batteries are likely to grow in excess of the market growth rate in the near term as its key markets — consumer electronics, hybrid vehicles, and telecom applications, are all expected to grow faster than GDP. Additionally, we believe this battery type is likely to cannibalize sales from NiCd batteries because of its non-toxic componentry and similar prices with NiCd. Longer term, however, we believe NiMH itself will see sales cannibalized significantly by Li-ion given higher energy density and potential for lower costs of Li-ion with scale. Key manufacturers of NiMH batteries for the hybrid market include Cobasys (ENER), Panasonic EV Batteries (Toyota), Sanyo, and SAFT. On the consumer battery side, Sanyo, Panasonic, Maxell, Gold Peak Industries and Yuasa manufacture NiMH batteries.

Hybrid Vehicle Market to Drive NiMH Battery Growth Near Term. As Chart 4 illustrates, the global hybrid vehicle market is expected to ramp significantly over the next 15 years, with hybrids capturing 5% market share in the U.S. early in the next decade. This expected explosion in growth is quite achievable in our opinion, due to higher gasoline costs and concerns about the environment. Today’s story, and the bulk of the near-term growth prospects in the industry, are in Ni-MH technology as the technology offers relatively high energy density and lower weight relative to lead-acid and NiCd batteries (see Chart 5). At the same time, production volumes have brought costs of battery packs down to reasonable levels and there are limited safety concerns related to Ni-MH — two key positives vs. Li-ion technology.

While enthusiastic on the NiMH in the near term, longer term we believe that Li-ion technology is likely to eclipse NiMH in hybrids (see below). While the rate of the technology shift is contentious, we have seen estimates for Li-ion market share in HEVs as high as 80% by 2017. If the safety profile of early Li-ion models is borne out, we do not believe these estimates are a stretch — despite the elongated planning processes of the large automakers.




Page 167 of 212

GLOBAL HYBRID VEHICLE GROWTH

Global Hybrid Demand Projections

0

2,000

4,000

6,000

8,000

10,000

2005 2010 2015 2020

Uni

ts(0

00's

)

OtherChinaJapanW. EuropeN.A.

Source: Freedonia

Lithium-ion and Lithium Metal

Lithium-based batteries are the newest commercial battery technology, first introduced by Sony in 1991. The initial application for these batteries was in consumer electronics as portable batteries for laptops and cell phones. Lithium-based batteries now serve the consumer electronic industry, military aircraft, space applications and potentially, the hybrid vehicle market. As depicted in Chart 5, these batteries have higher energy density than predecessor technology, which, along with their lower weight, have made them the technology of choice for the consumer electronics industry and a natural progression in the HEV market.

COMPARATIVE ANALYSIS OF BATTERY TECHNOLOGIES

Ni-MH

100 101 102 103 104

1000

100

10

1

64

2

64

2

64

2 1 h 0.1 h 36s 3.6s

Specific Power (W/kg)

Spec

ific

Ener

gy(W

h/kg

)

EV goalFuel Cells

HEV goal10 h

100 h

Acceleration

Ran

ge

Capacitors

Li-ion

IC Engine

Lead Acid

Source: LFEE (2007)



EXHIBIT 5:

EXHIBIT 6:


Page 168 of 212

While getting closer to ideal for performance metrics, there remain several issues holding back Li-ion technology including safety issues (thermal runaways in laptops led to recent recalls from Sony and Dell), calendar life, cycle life and cost shortcomings.

Thermal runaway has been the key issue that needs to be resolved for higher power applications, and particularly for hybrid vehicles. While fires in laptops are dangerous and have led to expensive recalls, similar issues, should they arise in vehicles, are likely to be deadly and a significant public safety hazard. The stability problems revolve around the presence of both a flammable electrolyte and unstable metal oxides in traditional Li-ion cells. Overheating, or in the case of a vehicle, a collision, could potentially lead to a chain reaction and fires in the battery pack.

Tackling this issue, along with calendar and cycle life issues has been at the core of development work for HEVs since the late 1990s and appears to be finally reaching a turning point. While concerns over stability remain, the issue appears to have been successfully addressed by the use of better control mechanisms and novel chemistries. Saft in particular has focused on the use of both a novel positive active material (Li[Ni,Co]O2), but also sophisticated control mechanisms designed to monitor chemical reactions in the cells. Other players such as A123 appear to have chosen more stable positive active material (LiFePO4), while newer technologies are focused on nano-structured cells and novel materials. While the technology remains a work in progress, current solutions appear to have addressed concerns sufficiently that the first widely commercialized vehicles with Li-ion technology are expected in 2009.

While much has been made of the cost disadvantage to Ni-MH (which remains significant given current volumes), higher production levels likely will bear out the lower fundamental cost of Li-ion cell production. Lower costs of Li-ion cells are driven both by their amenability to scale production and also due to the far lower metal content and fungible component materials in Li-ion cells.

NIMH AND LI-ION CELL COST COMPARISON

1.40

1.30

1.20

1.10

1.00

0.90

0.80

0.70

0.60

0.500 500 1,000 1,500 2,000 2,500 3,000

NiMHLi-Ion

Rel

ativ

eC

ellC

ost

HEV Volume (103)

NiMH vs. Li-Ion HEV Cell Cost/Volume Curve (50k-3M HEV/year)

1.40

1.30

1.20

1.10

1.00

0.90

0.80

0.70

0.60

0.500 500 1,000 1,500 2,000 2,500 3,000

NiMHLi-Ion NiMHLi-Ion

Rel

ativ

eC

ellC

ost

HEV Volume (103)

NiMH vs. Li-Ion HEV Cell Cost/Volume Curve (50k-3M HEV/year)

Source: LFEE – from an internal Ford study (2007)

As we remain in the early stages of commercial Li-ion technology development, there are many players currently developing Lithium-based batteries. Some of the players in this space are: China BAK Battery, Ener1, Japan Storage Battery, Matsushita, Panasonic EV Energy (Toyota), Sanyo, SAFT, Hitachi, Sony, and Valence Technologies. The space remains wide open given the early stage of technology development, but in HEV cells the early leaders appear to be Saft (which expects to see commercial production of its cells in a Mercedes in 2009), and A123 Systems (which is working on the GM E-FLEX system). Toyota/Panasonic, the early leader in Ni-MH hybrids, is also a key player in the space, but has decided to delay the rollout of Li-ion models until at least 2010.



EXHIBIT 7:


Page 169 of 212

EARLY LEADERS IN LI-ION BATTERY TECHNOLOGY FOR HYBRID VEHICLES

Battery Maker Integrator Technology Partners Stage of development

A123 Systems LiFePO4 Working on GM E-FLEX system Prototypes on E-FLEX being evaluatedHitachi GM - single model contract Hope to launch first Li-ion vehicle in 2010

LG Chem HyundaiWorking on GM E-FLEX system Prototypes on E-FLEX being evaluated

Mitsubishi/GS Yuasa Mitsubishi Peugeot Citroen - JV Hope to launch first Li-ion vehicle in 2010NEC Renault/Nissan - JV Starting manufacturing in 2009

Saft/Johnson Controls JCS Li[Ni,Co]O2Mercedes, second E.U. playerSAIC, Ford, Chrysler li-ion development projects

To begin commercial production of Mercedes S Class in 2009.

Sanyo Believed to be non-phosphate VW - not JV, single model deal To start building commercial capacity in 2009

Toyota/Panasonic Toyota Believed to be Li[Ni,Co]O2 Toyota - JV Have delayed Li-ion based Prius until 2010

Source: Company Documents, Jefferies Intl. Ltd.

Outlook. We expect the lithium based batteries to grow faster than the consumer markets and incrementally take share from Ni-MH. While expecting Li-ion to take significant share in the HEV market given the inherent benefits of the technology, we expect the transition to be gradual, pending greater certainty on the stability issue of the cells and the ramp in scale which will make Li-ion cost competitive.

Ultracapacitors/Supercapacitors

Over the previous couple of years, the industry has grown 75%–100% annually, which we expect to continue at least through 2010 and likely for a significantly longer period of time. We believe the ultracapacitor provides solutions for engineers that were not possible even five years ago and we expect that customer acceptance of the product will continue to accelerate post 2010. We expect that as ultracapacitors become more widely known, additional applications will drive the growth longer term, which will increase the addressable market. As Table 2 demonstrates, ultracapacitors provide a better blend of power delivery and energy storage than either a pure capacitor or a battery. The benefits of ultracapacitors versus batteries are detailed in Table 2, but in our opinion, the two most important characteristics are the fast charge/discharge cycle (half a second) versus 8 hours for a lead acid battery and the nearly unlimited charge/discharge cycles (500K cycles versus <1000 for lead acid battery). The main disadvantages of ultracapacitors are that they cannot hold as much energy as a similarly sized battery and to a lesser extent, price. We believe that this indicates that both are likely to be used for specific applications. But in the case of ultracapacitors, the growth will be significantly higher because the base is low and they are creating new applications. The main competitors in this space are Maxwell Technologies, Montena, Ness Corporation, and Panasonic EV Energy.

ULTRACAPACITOR CHARACTERISTICS

Ultracapacitor vs. Battery100x more instantaneous power than batterieslower weight vs. electrical energy storeddeeper dischargelower wasted energy (less heat)infinite charge/discharge cycles (extended Useful life)better temp. variance (-40C to 65C)environmentally friendlyBetter Reliability - No Moving Parts or Chemical ReactionU/C allow for smaller engines or batteries due to peak power requirements handled by U/C'sLong Shelf LifeMaintenance Free

Source: Jefferies Intl. Ltd.



EXHIBIT 8:

EXHIBIT 9:


Page 170 of 212


ULTRACAPACITOR VERSUS BATTERY COMPARISON

0,01

0,1

1

10

100

1,000

10 100 1,000 10,000

BOOSTCAPUltracapacitor

Lead Acid Battery

Ni -Cad Battery

Lithium Battery

Conventional Capacitor

10% the Energy of a Battery

10 hours

1 hour1 second

.03 seconds

SpeedPower Density/[W/kg]

Qua

ntity

Ener

gyD

ensi

ty/[W

h/kg

]

0,01

0,1

1

10

100

1,000

10 100 1,000 10,000

BOOSTCAPUltracapacitorBOOSTCAP

Ultracapacitor

Lead Acid Battery

Ni -Cad BatteryNi -Cad Battery

Lithium BatteryLithium Battery

Conventional Capacitor

10% the Energy of a Battery

10 hours

1 hour1 second

.03 seconds

SpeedPower Density/[W/kg]

Qua

ntity

Ener

gyD

ensi

ty/[W

h/kg

]

Source: Maxwell Technologies, Jefferies & Company, Inc.

Top-Down Analysis Implies a Potential $12B Market in 2010. As it matures, the industry has been able to reduce costs to the end customer, which increases the potential addressable market significantly. In general, for every 50% reduction in cost to the end customer, the potential volume increases by a factor of 10. A simple example helps to demonstrate the point. A typical hybrid city bus costs around $300K today. To replace the batteries in a hybrid bus with ultracapacitors requires around 300 large cell (2700 farad) ultracapacitors. At $400 a piece in 1999, this equated to $120,000 for just the ultracapacitors alone, and thus ultracapacitors were not a viable option for the bus company. But today, at $25 each, the cost is only around $7,500, or 2.5% of the total bus cost. In a similar fashion, the cost of ultracapacitors is already low enough for the hybrid and luxury personal vehicle markets, but perhaps not quite cheap enough for mainstream autos. However, with another expected 50% reduction in expected end costs, we expect that ultracapacitors likely will become a valid alternative for automakers. Of course, given long-term design requirements of as long as seven years, it will likely take several years before ultracaps can achieve full market penetration. Importantly, while the auto industry is a significant long-term growth driver, industrial, telecom, and consumer electronics will likely drive growth in the near term.

Other Technologies – A wide range of energy storage solutions are emerging to meet specific niche application requirements. These include alkaline-based batteries, flywheels, lithium-polymer combination cells, Nickel hydrogen, and Zinc-air cells. Given the wide range of end user needs and specifications, the various emerging battery technologies have further fragmented and specialized the battery market, but in our opinion, are not likely to displace the above-mentioned technologies, except for customized applications. However, we believe the industry is shifting toward batteries with higher-energy densities. As these technologies ramp production and reduce costs, we believe this shift could accelerate.

Conclusions

We expect overall demand for batteries to grow above global GDP as increasing globalization and rapid growth in undeveloped countries is likely to drive increased demand growth. We believe that the trend is toward those batteries with higher energy densities; however, the customer will not pay for higher energy density unless the application requires it. Markets such as forklifts for which lead acid batteries are well suited are likely to stick to existing technologies. In markets such as telecom reserve power, we are likely to continue to see only limited conversion to newer technologies because back-up power is rarely used (only during power outages) and lead acid remains the cheapest solution where maintenance access poses minimal problems. In contrast, hybrid vehicles require greater energy density to power the drivetrain and existing Ni-MH technology leaves lots to be desired in terms of cost and performance. Therefore, in that application, we believe auto OEMs will pay a higher price for newer technologies which can improve existing performance. In total, we believe that Li-Ion batteries and Ultracapacitors are likely to grow faster than the overall market, while lead-acid (automotive applications) and NiCd batteries are likely to grow more slowly than the overall market.


EXHIBIT 10:


Page 171 of 212





EventThis report details the benefits and applications that the fuel cellindustry could achieve, as well as the principal challenges facingthe industry and a look at individual fuel cell technologies.

Key Points• Cost. The cost of fuel cells is significantly higher than both

traditional forms of generation and power storage. We believethat prices for most technologies must be reduced significantlybefore fuel cells can be rolled out on a commercial scale foreither generation or storage applications.

• Durability. Fuel cells tend to be very fragile and unproven whilethe existing technology they often seek to displace is wellknown, battle tested, and quite durable. There is a wide rangeof durability issues that the industry faces across the entirespectrum of technologies and applications.

• Infrastructure. The most commonly known form of fuel cell is ahydrogen-powered PEM fuel cell, often used in transportation inlieu of an internal combustion engine. Unfortunately, hydrogenis not as ubiquitous as gasoline and faces significantinfrastructure hurdles due to the incompatibility of much of theexisting gas pipelines and hydrogen.

• Environmental benefits. While the previous points illustratethe challenges, it is time to recognize the potential benefits.Fuel cells generally have lower emissions and higher efficiencythan the technologies they displace. This is true in virtually allapplications.

• First adopters. We believe that early adopters could bestationary generation and portable military/security users. Whilethese are radically different applications, the specifictechnologies offer tremendous potential to improve on existingsolutions.

October 6, 2008

Clean TechnologyEnergy Storage - Fuel Cells

Clean TechnologyFuel Cells Primer

Investment SummaryWe believe that fuel cells offer tremendous potential forconversion of stored energy in both stationary and mobileapplications. However, the high costs and performance issueswill remain significant barriers to mass deployment for severalyears.




Executive Summary

The basis for modern fuel cell technology emerged over 150 years ago when Welsh lawyer turned scientist William Robert Grove garnered international acclaim for his “Grove cell.” Fast forwarding to the late 20th century, fuel cells re-emerged in the public consciousness as a potential solution for both higher energy prices and environmentally damaging emissions. The idea of an electric car propelled by hydrogen powered fuel cells emitting nothing more than steam caught the public’s eye and attracted hundreds of millions of dollars in investment from both private investors and governments.

A workable fuel cell solution has many attractive benefits. Looking just at the transport sector, enthusiasts dream of replacing gasoline fired engines with quieter and cleaner electric engines. The switch from fossil fuels to hydrogen powered fuel cells could virtually eliminate local emissions from the transport sector with even greater savings potential for stationary power. As the chart on the right shows, road transport represented 21% of carbon emissions in the UK in 2004 with over double that amount from electricity usage. We believe that other industrialized countries have similar emission levels. Any significant reduction in emissions from using fuel cells in vehicles, buildings, and industry could have a significant impact on the build-up of greenhouse gases and reduce concerns over global warming. Another benefit would be the potential reduction in dependence on oil. Energy security is an increasingly important issue for governments around the world. To this end, many governments have set aside significant sums for development of fuel cells.

EXHIBIT 1: CARBON EMISSION PROFILE

However, despite all of the obvious attractions of fuel cell potential, the idea of a fuel cell or hydrogen economy has not yet been turned into reality. Although fuel cells have been selling into niche markets on a commercial basis for some time, mass market applications have yet to take off. Given that we believe the benefits of a hydrogen or fuel cell economy are relatively clear and straightforward, we have chosen to focus on some of the challenges the fuel cell industry faces in commercializing the product. However, first we offer a quick overview of fuel cell technology and types of fuel cells.

Fuel Cell Technology Primer

While we do not intend to offer a scientific dissertation on the inner workings of a fuel cell, a quick summary of fuel cell technology basics may be in order. While different types of fuel cells have different operating characteristics, in general a fuel cell operates in the following manner.

Hydrogen (H2) molecules enter the fuel cell at the anode where a chemical reaction strips them of their electrons. The ionized (carrying a positive electrical charge) hydrogen atoms pass through the electrolyte membrane. The negatively charged electrons pass through a wire outside the fuel cell to create electrical current. The electrons and protons mix with air on the cathode to complete the reaction and create water as an exhaust. See below.

Energy Storage - Fuel Cells

UK CO2 Emissions by Source (2004)

Road Transport

21%

Other8%

Residential16%

Other Industries

18%Energy

Industries37%

Source: DEFRA



EXHIBIT 2: FUEL CELL DIAGRAM (FOR PEM TYPE FUEL CELL)

Source: Fuel Cells 2000

Fuel Cell Applications

As we mentioned in the Executive Summary, hydrogen powered PEM fuel cells remain a compelling and probably the most well know application for fuel cells. However, there are a vast range of applications that can be met with a variety of fuel cells types (we will review the various fuel cell technologies in the following section). While we cannot say if fuel cells will ever be deployed in the following applications, we present an abbreviated sample of possibilities:

� Car engine replacement. The application that immediately springs to mind when investors and the general public think of fuel cells. A hydrogen-fuelled electric engine could eliminate harmful tailpipe emissions and reduce dependence on imported oil. However, there are significant cost, durability, and fueling issues that must be addressed.

� On-board electronics. Nearly 20% of the U.S. truck fleet fuel consumption is used when trucks are idled but electric power must be maintained. For example, refrigerated trucks must keep their engine running lest the refrigeration unit lose power and the perishables are lost. A fuel cell stack attached to a natural gas or propane tank could meet this demand and reduce idling losses. However, cost and durability have been a key barrier here.

� Military/Security. The modern soldier carries far more electronics than his predecessors and the resulting increase in power needs has led to heavier and heavier rucksacks as the soldiers are forced to carry extra batteries. Fuel cells could replace batteries in certain applications although cost and performance remain key concerns.

� Stand-alone generation. Diesel generators are the solution of choice for many stand-alone applications (telecom relay stations, construction sites, USP, etc.). A low cost fuel cell stack could dislodge the inefficient diesel generator. Again, cost and durability are key issues.

� Residential heat and power. As certain types of fuel cells generate both heat and electricity using natural gas as a fuel, it is potentially possible to provide heat and power to the home using a fuel cell stack. Cost, durability, and performance have been key barriers.

� Laptop battery augmentation / replacement. A lightweight and easily rechargeable fuel cell could allow a laptop to run for days rather than hours without seeing an electrical outlet. However, cost and competition from alternative solutions remain key concerns.



Page 175 of 212


Types of Fuel Cells

The term “fuel cell” has often been associated with low temperature hydrogen powered cells commonly known as PEM (Proton Exchange Membrane) fuel cells. However, fuel cells come in many different forms, and can use a range of fuels and have widely differing thermal and performance characteristics.

EXHIBIT 3: FUEL CELL TYPES

Type Comments PEM Require a polymer electrolyte membrane and metallic catalysts. PEM fuel cells require

ultra-pure hydrogen as an input with this “fuel” needing to be created / synthesized elsewhere. Given their thermal characteristics (operating typically at 0-100°C), PEM fuel cells are the preferred solution for fuel cell powered vehicles.

Direct Methanol Direct methanol fuel cells are similar to PEM fuel cells except that they convert liquid methanol, rather than hydrogen gas, into electricity. This technology is being actively developed by many leading Japanese corporations as a replacement for high capacity batteries for laptops and other relatively demanding portable devices.

Solid Oxide Solid oxide fuel cells use hard ceramic electrolytes and operate typically at very high temperatures (800-1000°) and with superior efficiency than lower temperature PEM cells. Ceramic based solid oxide fuel cells do not require pure hydrogen and can operate with hydrocarbons (e.g., natural gas) as a fuel. Given their operating characteristics, solid oxide fuel cells are generally associated with fixed combined heat and power (CHP) generation projects.

Molten Carbonate Molten carbonate fuel cells share many similar characteristics with solid oxide fuel cells with the major exception that they are based on molten carbonate salts and have significantly more complex system design limiting their use to very large scale plant (1>MW).

Alkali The fuel cell of choice for early NASA projects. Alkali based fuel cells do not require membranes but still require catalysts and operate at 150-200°C. Pure hydrogen is also a requirement for alkali.

Source: Jefferies equity research

Each type of fuel cell faces its own unique set of challenges. In the following sections, we have attempted to highlight the primary issues that fuel cell developers confront.

Cost

Exorbitant cost has been the principal issue for the fuel cell industry. According to the U.S. Department of Energy (DoE), fuel cells currently cost approximately $3000 per kW output, although many industry experts believe that this cost is too low by anywhere from 30%–70%. The DoE believes that fuel cells cannot compete with traditional forms of generation unless the cost per kW output is $1000. The most significant costs of a fuel cell have been:

� Materials. Industry standard polymers for PEM fuel cells currently cost approximately $500/m2 while specialty ceramic powders used in solid oxide fuel cells can be very expensive as well.

� Fuel. While hydrogen is plentiful it is always locked up in the form of water, hydrocarbons, etc. Pure hydrogen is not found anywhere on earth, forcing many fuel cell types to rely on synthesizing hydrogen at great financial cost and environmental damage.

� Catalyst. Low temperature fuel cells require precious metal catalysts to ensure a successful electrochemical reaction. Currently, the industry standard is platinum, which has increased 26% in the last year to almost $1100/ounce. While other catalysts are under development, platinum continues to be the main catalyst.

� Assembly. Fuel cells require very high levels of engineering to produce a successful product. While economies of scale will undoubtedly reduce production costs, there are a number of economic issues that must be addressed before mass production becomes a reality.



Page 176 of 212


Fuel cell developers are constantly seeking to address these issues, and efforts are under way at various companies to reduce the cost of fuel cells. Furthermore, certain types of fuel cells are exempt from some of the cost issues mentioned above. For example, some higher temperature and alcohol fuel cells do not require pure hydrogen which obviates the need for expensive fuel costs. Also, high temperature fuel cells do not require catalysts. The disadvantage to high temperature fuel cells is that they are significantly larger and not suitable for transportation applications.

Durability

There is no point in commercializing a fuel cell if its effective productive life is too short. Fuel cells must be able to survive and function at high levels of efficiency in real world conditions. As is the case with the cost issue, the durability issue differs based on fuel cell types.

� PEM fuel cells have two primary durability concerns. First, the membrane must be precisely hydrated in order to operate and failures in hydration can reduce operational life. If it gets too wet the chemical reaction slows while if it dries out it can crack. Second, catalysts in a fuel cell can also be poisoned by impurities in the hydrogen fuel leading to decaying performance, or prohibitively expensive purification equipment.

� Solid Oxide fuel cells can suffer from thermal cycle-ability issues. In other words, they cannot typically be turned on and off. Given the high operating temperatures, the ceramic will expand and contract as the temperature rises and falls leading to failure of the fuel cell. Additionally, solid oxide fuel cells are significantly heavier than PEM fuel cells.

While many investors may believe that durability and cost are the same issue, we believe that they should be addressed separately. While durability undoubtedly impacts costs, we feel the practical affect of the durability issue is equally important. Just imagine a fuel cell powered car that needs a new engine every 6–12 months. Even if cost is not an issue, this is not an attractive consumer application.

Fuel Demands

Fuel cells and hydrogen have long been inextricably linked in investor’s minds, with some justification. However, not all fuel cell types require pure hydrogen. This is important because sourcing hydrogen for fuel cells, or for any other reason, creates several major challenges.

� Pure hydrogen does not exist on earth and thus must be manufactured. The most common and by far the most economic method is to reform natural gas in order to remove carbon, methane, and other molecules to be left with pure hydrogen. However, this is hardly ideal as a fossil fuel will remain as the feedstock for the hydrogen economy, reforming releases harmful emissions, and much of the original energy content in the fuel is lost in the process. The other more expensive method is electrolysis where water is broken down to the molecular level to create hydrogen and oxygen. However, as an electrolyzer is essentially a fuel cell in reverse, the cost and durability issues previously mentioned persist.

� Electricity is also needed in large quantities to power an electrolyzer. The vision of renewable energy powering an electrolyzer to produce hydrogen out of water to be used to power an emission free economy is very appealing in principle. However, we suffer from a shortage of renewable energy generation, and it makes little sense to divert the limited supply of renewable energy towards the hydrogen economy as the diverted renewable energy will have to be replaced with dirtier fossil fuel generation. In addition, due to the high cost of renewable power, “clean” hydrogen produced via this route is inherently expensive.

� Transporting hydrogen cannot be accomplished using existing hydrocarbon distribution infrastructure, as differences in properties between pure hydrogen and these fuels can lead to leakage as well as damage to the pipelines themselves, tanker trucks require major modification, and gasoline tanks are useless

� Storing hydrogen gas is problematic for a couple of key reasons. Hydrogen’s very low energy density means that much more gas (approximately 3X by volume when compared to natural gas and approximately 100X compared to gasoline) is needed to provide the same energy output. Therefore, storing enough hydrogen either requires large, extremely high pressure tanks, cryogenic liquefied H2, or storage in solid metal - each of these solutions being inherently problematic. This leads us to the second main issue—reactivity. Pure hydrogen is more reactive in the presence of oxygen than natural gas; this means it burns more easily and is inherently more dangerous.



Page 177 of 212


Hydrogen Fuel

Hydrogen does have one key benefit, it does not have to be utilized as an input for a fuel cell to be effective. With a minimum of modifications, hydrogen can be used in most internal combustion engines in lieu of gasoline. Pure hydrogen has lower harmful carbon emissions when burned in an internal combustion engine. However, in order for hydrogen to be used as a gasoline substitute you need—hydrogen. In addition to all of the cost, durability, and storage issues we mentioned previously, hydrogen refueling opportunities would have to be either ubiquitous for mass use or at a central location for fleet use. Additionally, the low energy density of hydrogen gas reduces travel radius while storing a highly pressurized, reactive gas in a moving car could prove dangerous in the case of an accident. However, if these critical issues can be overcome, the potential for hydrogen to displace gasoline could provide a significant opportunity.

Conclusion

Taken as a whole, the fuel cell industry still faces a long road of technical development and cost reduction before the product is likely to achieve mainstream acceptance. Given the volatility in the space and previous hype that surrounds the stocks, we would recommend a cautious approach to investing in the sector. However, within this broad category, we believe that individual technologies and applications could offer significant potential in terms of improving energy efficiency, emissions reductions, or both. The key will be to recognize the potential benefits of the specific fuel cell application, the challenges that must be overcome, and the potential size of the target market and estimated speed of adoption of the new technology.

Fuel Cell Companies

The universe of companies involved in fuel cell research and development is quite diverse. It ranges from small private companies with no commercial products to large capitalization, publicly traded companies developing commercial products. The main companies in the space include the following:

• Ballard Power Systems (NASDAQ: BLDP, NC) • Ceres Power (AIM: CWR, Buy, 454p PT) • Distributed Energy (NASDAQ: PLUG, NC) • Energy Conversion Devices (NASDAQ: ENER, Buy, $96 PT) • FuelCell Energy (NASDAQ: FCEL, NC) • Hydrogenics Corporation (NASDAQ: HYGS, NC) • International Fuel Cells, a United Technologies Company (NYSE: UTX, Buy, $85 PT) • ITM Power (AIM: ITM, NC) • Mechanical Technology (NASDAQ: MKTY, NC) • Millenium Cell (NASDAQ: MCEL, NC) • Plug Power (NASDAQ: PLUG, NC) • Angstrom Power (Private) • Jadoo (Private) • Nuvera Fuel Cells (Private) • Protonex (Private) • Ultracell (Private)



Page 178 of 212

EventJefferies Research has collaborated to provide an overview of theeconomics of nuclear power, with a focus on greenfield economicsand the need for subsidies to generate an adequate return on capital.

Key Points• Nuclear investments require subsidies. In the absence of

sizeable subsidies by the federal government, Jefferies believes theeconomics do not justify building merchant nuclear generation inthe U.S. We estimate the U.S. government's current $18.5bn loanguaranty program could support 3-4 of the 19 new nuclear powerplants that have been proposed.

• Costs can be significant. With costs for new nuclear generationcurrently ranging from $2,400 to $4,540 per kilowatt (kW), weestimate that power prices of at least $100/MWh would be requiredto economically build new nuclear generation without subsidies.Taking into account the government subsidies that are in place,however, we estimate the economics on new merchant capacityshould work. Moreover, a significant price for carbon or stateincentives could improve plant economics.

• Fuel costs can have an impact. Fuel costs are often viewed as asecondary consideration given the large capital costs involved in anew nuclear plant. We believe this makes sense only on theassumption of significant federal subsidies. For example, assuminga uranium enrichment company such as USEC (USU, $3.44, Buy)aims for a 10% ROIC on its investments in new capacity withoutany government subsidy, we estimate fuel costs could cut the IRRon a new nuclear plant by 15-40bps. In effect, a $0.25/MMBTUincrease in fuel prices has the same impact on IRRs as a $50/kWchange in EPC costs or a 360bps change in interest rates.

October 10, 2008

Nuclear Primer

Investment SummaryIn the ongoing debate over the environmental footprint of nuclearpower and the political will to build new plants, underlying economicsare often a footnote. Our analysis of existing government subsidiesand construction costs suggests the economics for new merchantcapacity can work.

Debra E. Bromberg(212) 284-2452, [email protected]



Debra E. Bromberg , [email protected], (212) 284-2452

Merchant Nuclear Generation: Do the Economics Make Sense?

Our analysis of greenfield nuclear plant economics suggests construction of new nuclear generation in the U.S. hinges on government subsidies. In the absence of sizeable subsidies by the federal government, Jefferies believes the economics do not currently favor the building of new nuclear generation in the U.S. Based on assumed average EPC (engineering, procurement and construction) costs for new nuclear generation of approximately $3,500 per kilowatt (kW) and all-in costs of $5,225 per kW, we estimate that a power price of at least $100 per megawatt-hour (MWh) would be required to build new merchant nuclear generation in the absence of federal subsidies. When factoring in these government incentives, however, the economics should work. We have focused our analysis on merchant generation because regulated utility nuclear generation costs would be even higher due to an assumed capital structure that would be less leveraged.

According to the Nuclear Energy Institute (NEI), the first of the next generation of nuclear plants in the US could be on-line as early as 2015. The first few plants are expected to be built in the Southeast (see Exhibit 1), and whether merchant or regulated, they will receive federal government incentives in the form of loan guarantees and/or production tax credits (PTC). As of October 1, 2008, the U.S. Department of Energy (DOE) has received 19 applications for federal loan guarantees totaling $122 billion. Although the amount of available loan guarantees could eventually be increased, congress has authorized $18.5 billion, which would cover about 3-4 of the first plants built. PTCs are available to the first 6,000 MW and are pro-rated and capped per plant.

Source: Reprinted from SNL Energy Power Daily

EXHIBIT 1: STATUS OF PROPOSED U.S. NUCLEAR POWER PLANTS


Page 180 of 212

Our base case analysis (see Exhibit 2), which is based on IRR, assumes EPC costs of $3,500 per kW (for a 1,400 MW plant), all-in costs of $5,116 per kW, a 20% equity ratio, four-year construction period and 90% capacity factor. We note that the average capacity factor for the industry in 2007 (based on 104 plants operating in the U.S.) is estimated at 91.8% according to the Energy Information Administration. Our IRR hurdle rate of 4.1% assumes that 100% of debt is government guaranteed with an interest rate of 4.5%. The IRR achieved in our analysis is 4.4% and we estimate that to achieve an IRR that is double our hurdle rate (or 8.2%), a power price of $119 per MWh is required.

For all-in costs, we have assumed a 10% contingency factor, 2.5% inflation rate, owner’s costs of $286 per kW (which includes $100 million of development costs) and a credit subsidy cost of 5%. A number of our base-case assumptions were taken from an August 2008 white paper by NEI titled “The Cost of New Generating Capacity in Perspective,” although our fuel cost assumption is somewhat lower, at $0.50 per MMBtu, our cost per kW is at the low end of NEI’s range, and our O&M assumption is slightly higher at $10.00 per MWh. Our analysis also assumes an average price per MWh of $70.00—which is the average of NEI’s assumed year-one price in its analysis--throughout the plant’s life.

We estimate that a $100 per kW change in EPC costs affects the IRR in our analysis by 49 basis points, a $0.25 per MMBtu change in fuel price has a 25 basis point impact and a 100bps change in interest rates has a 7 basis point impact.

The analysis in Exhibit 3 incorporates the same assumptions as in our base case scenario, except that we assume a higher power price (at least $100 per MWh) is necessary to achieve an IRR that is slightly above our 5.8% hurdle rate. This scenario assumes no federal loan guarantees or PTCs and an 8.0% cost of debt.



Page 181 of 212

New Merchant Nuclear 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021($ MM)Power Price per MWh 70.00$ 70.00$ 70.00$ 70.00$ 70.00$

Cost per kW EPC costs per kW 3,500$ 5,116$ 5,116$ 5,116$ 5,116$ 5,116$Capital (all-in) kW (MM) 1.400 7,162 7,019 6,852 6,660 6,444% Equity 20% 20% 21% 20% 20%Assumed Debt EPC costs 4,900$ 5,730 5,586 5,443 5,300 5,157Assumed Equity EPC inflation factor (2.5%/yr) 509 1,432 1,432 1,409 1,360 1,287Assumed Debt Amortization 10% contingency factor 490 (143) (143) (143) (143) (143)

owner's cost 400Assumed Depreciable Plant Life overnight costs 6,299$ 36 36 36 36 36

Depreciable life over-ride credit subsidy cost 252 35.7 35.7 35.7 35.7 35.7Interest Rate % capitalized interest 611 5% 5% 5% 5% 5%

Interest rate over-ride all-in costs 7,162$ 4.5% 4.5% 4.5% 4.5% 4.5%Tax Rate % 38% 38% 38% 38% 38%

Tax rate over-ride Jefferies all-in cost per kW 5,116$

Megawatt (000's) 1.400 1.400 1.400 1.400 1.400Assumed Capacity Factor 90% 90% 90% 90% 90%Megawatt-hours 11.038 11.038 11.038 11.038 11.038Fuel Cost per MWh 5.00$ 5.00$ 5.00$ 5.00$ 5.00$Heat Rate 10.000 10.000 10.000 10.000 10.000$ Cost per MMBtu 0.50$ 0.50$ 0.50$ 0.50$ 0.50$

Margin 717 717 717 717 717O&M per MWh 10.00$ 10.25$ 10.51$ 10.77$ 11.04$Operating & Maintenance Costs (110) (113) (116) (119) (122)O&M per kW-year 78.84$ 80.81$ 82.83$ 84.90$ 87.02$Depreciation (175) (175) (175) (175) (175)Property Taxes (54) (54) (54) (54) (54)Operating Income 379 376 373 370 367

Interest Expense (258) (251) (245) (238) (232)Amortization of Debt Costs (22) (22) (22) (22) (22)Pre-tax Income 99 103 106 110 113Income Tax (38) (39) (40) (42) (43)PTCs 175 175 175 175 175Net Income 236 239 241 243 245

Cash Flow 290 292 294 296 298Dividend 236 263 289 316 368ROE 17% 17% 17% 18% 19%

EBIT 379$ 376$ 373$ 370$ 367$Less income taxes (235) (233) (231) (229) (228)Earnings before interest and after tax 144 143 142 141 139Plus depreciation & amortization 196 196 196 196 196Less credit subsidy costs (252)Less development costs (50) (50)Plus deferred taxes 92 92 92 92 92Plus PTC's 175 175 175 175 175Less capital spending (6,199) (1,550) (1,550) (1,550) (1,550) (14) (14) (14) (14) (14)

Unlevered Free Cash Flow (1,852) (1,600) (1,550) (1,550) 594$ 593$ 592$ 590$ 589$

Source: NEI and Jefferies & Company, Inc. estimates

EXHIBIT 2: BASE CASE MERCHANT NUCLEAR GENERATION: FIRST 5 YEARS EARNINGS & FCF PROJECTIONS



Page 182 of 212

New Merchant Nuclear 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021($ MM)Market Price per MWh 100.00$ 100.00$ 100.00$ 100.00$ 100.00$

Cost per kW EPC costs per kW 3,500$ 5,225$ 5,225$ 5,225$ 5,225$ 5,225$Capital (all-in) kW (MM) 1.400 7,315 7,169 7,008 6,833 6,642% Equity 20% 20% 21% 21% 21%Assumed Debt EPC costs 4,900$ 5,852 5,706 5,559 5,413 5,267Assumed Equity EPC inflation factor (2.5%/yr) 509 1,463 1,463 1,449 1,420 1,375Assumed Debt Amortization 10% contingency factor 490 (146) (146) (146) (146) (146)

owner's cost 400Assumed Depreciable Plant Life overnight costs 6,299$ 36 36 36 36 36

Depreciable life over-ride credit subsidy cost - 35.7 35.7 35.7 35.7 35.7Interest Rate % capitalized interest 1,016 8% 8% 8% 8% 8%

Interest rate over-ride all-in costs 7,315$ 8.0% 8.0% 8.0% 8.0% 8.0%Tax Rate % 38% 38% 38% 38% 38%

Tax rate over-ride Jefferies all-in cost per kW* 5,225$NEI avg. all-in cost per kW 5,725$

Megawatt (000's) 1.400 1.400 1.400 1.400 1.400Assumed Capacity Factor 90% 90% 90% 90% 90%Megawatt-hours 11.038 11.038 11.038 11.038 11.038Fuel Cost per MWh 5.00$ 5.00$ 5.00$ 5.00$ 5.00$Heat Rate 10.000 10.000 10.000 10.000 10.000$ Cost per MMBtu 0.50$ 0.50$ 0.50$ 0.50$ 0.50$

Margin 1049 1049 1049 1049 1049O&M per MWh 10.00$ 10.25$ 10.51$ 10.77$ 11.04$Operating & Maintenance Costs (110) (113) (116) (119) (122)O&M per kW-year 78.84$ 80.81$ 82.83$ 84.90$ 87.02$Depreciation (175) (175) (175) (175) (175)Property Taxes (55) (55) (55) (55) (55)Operating Income 709 706 703 700 697

Interest Expense (468) (456) (445) (433) (421)Amortization of Debt Costs (25) (25) (25) (25) (25)Pre-tax Income 215 224 233 242 250Income Tax (82) (85) (88) (92) (95)PTCs 0 0 0 0 0Net Income

133 139 144 150 155Cash Flow 187 193 198 204 209Dividend 133 153 173 195 233

ROE 9% 9% 10% 11% 11%

EBIT 709$ 706$ 703$ 700$ 697$Less income taxes (439) (438) (436) (434) (432)Earnings before interest and after tax 269 268 267 266 265Plus depreciation & amortization 200 200 200 200 200Less credit subsidy costs 0Less development costs (50) (50)Plus deferred taxes 92 92 92 92 92Plus PTC's 0 0 0 0 0Less capital spending (6,199) (1,550) (1,550) (1,550) (1,550) (14) (14) (14) (14) (14)

Unlevered Free Cash Flow (1,600) (1,600) (1,550) (1,550) 548$ 547$ 546$ 545$ 543$

Source: NEI and Jefferies & Company, Inc. estimates

NEI estimates that overnight costs for new nuclear generating capacity could range from $2,400-$4,540 per kW. Overnight costs include engineering, construction and procurement (EPC) costs as well as owner’s costs, such as infrastructure (e.g., transmission upgrades and cooling towers) and permitting and development costs. NEI attributes the wide cost range to uncertainty about escalation of commodity prices and wages, the inability to obtain a precise price because design work is not complete and early estimates not fully including all costs involved in construction.

NEI also estimates that the first-year busbar (all-in) cost for a new regulated nuclear plant ranges from $96.90-$118.80 per megawatt-hour (MWh) and the levelized cost is $73.60-$87.70 per Mwh (in 2007 dollars). NEI also estimates that a cash return on CWIP (construction work in progress) could reduce the first-year busbar cost by 20-30%.

These estimates assume an EPC cost of $3,500-$4,500 per kW and total cost of $4,351-$5,473 per kW. Total costs include overnight costs plus all capital costs, contingencies and financing costs. Other assumptions include 50%/50% debt/equity, $9.50 per MWh O&M cost, $7.50 per MWh fuel cost, 6.0% interest rate for commercial debt and a return on CWIP. This compares with NEI’s estimated first-year busbar cost for a new merchant nuclear plant of $64.40-$75.80 per MWh and a total cost of $5,071-$6,378 per kW. Other assumptions include 80%/20%

EXHIBIT 3: MERCHANT NUCLEAR GENERATION (NO SUBSIDIES): FIRST 5 YEARS EARNINGS & FCF PROJECTIONS



Page 183 of 212

debt/equity supported by a federal loan guarantee, 4.5% rate on government-guaranteed debt and a 5% loan guarantee cost.

We would agree with NEI’s conclusion that in the absence of a significant price for carbon, loan guarantees for merchant nuclear and supportive state policies (such as a return on CWIP for regulated plants) are essential for new nuclear plants to be financed and constructed competitively.

Government Incentives

Under a solicitation for nuclear power facility projects, the DOE is making available up to $18.5 billion in loan guarantees under Title XVII of the Energy Policy Act of 2005. Although under the terms of the Act the availability of the loan guarantee authority under the fiscal year 2008 Appropriations Act expires September 30, 2009, the DOE has requested in its FY2009 budget that the authorization be extended through September 30, 2011. Debt guaranteed by the DOE under Title XVII is limited to no more than 80% of total project costs.

Two other federal incentives for new nuclear included in the Energy Policy Act are production tax credits and risk insurance. The PTCs are available to 6,000 MW of new nuclear capacity at $18.00 per MWh for the first eight years of operation (maximum tax credit allowed per plant is $125 million per 1,000 MW annually). To be eligible for the PTCs, the COL license application needs to be submitted to the NRC by December 31, 2008, plant construction must start by January 1, 2014 and the plant must be operating by January 1, 2021. Facilities that qualify for the PTCs will receive them on a pro rata basis according to plant size.

Risk insurance (referred to as standby support) is available to the first six nuclear plants and covers licensing and litigation risk. It applies only to delays that are caused by factors that are beyond a company’s control and can be applied only to debt service. The first two plants constructed qualify for $500 million of standby support coverage and the next four plants can each receive up to $250 million of coverage. Delay coverage begins six months into a covered delay and covers 50% of eligible costs. The DOE estimates that the first conditional agreement could be signed as early as this December, with the first standby support contract signed in late 2010.

U.S. Representative Gresham Barrett has proposed legislation to expand certain incentives under the Energy Policy Act, such as broadening standby support and allowing public power entities that have partnered with private companies (i.e., to develop nuclear power plants) to take full advantage of nuclear tax credits.

In addition to federal incentives, a number of states have passed legislation or implemented policies that support new nuclear construction, such as allowing a return on CWIP, a prudence determination prior to construction (with periodic reviews during construction) and property tax incentives. The first two can substantially reduce cost recovery risk, while the first one (a return on CWIP) can also mitigate rate shock associated with a regulated plant following commercial operation.



Page 184 of 212

Bottom Line: Government Subsidies Matter

The following two tables are intended to provide a simple snapshot of year-one power prices necessary to achieve a 10% ROE for a 1,000 MW greenfield gas-fired CCGT, merchant nuclear plant and wind power project (Exhibit 4) and the impact on year-one ROEs of applying the same prices but excluding federal subsidies (i.e., for nuclear and wind). We note that ROEs for nuclear and wind drop off significantly in year nine and year 11, respectively, following expiration of PTCs.

We believe that to adequately evaluate the economics of any project, one needs to estimate the cash flows generated over the project’s life, factoring in, for example, the finite life of certain federal subsidies like PTCs and tax benefits due to accelerated depreciation (e.g., MACRS). Other factors to consider include the potential for capacity payments in particular markets (e.g., NEPOOL or PJM), the impact on power prices of regional transmission constraints (e.g., wind power located in western Texas), state incentives (e.g., property taxes), regional gas price basis differentials, local average wind capacity factors, and the impact on market power prices of existing (e.g., RGGI in the Northeast) or potential federal carbon legislation. We recognize the limitations of this “snapshot” analysis and note that we have more detailed power plant models available.

Scenario 1: Assumed Price per MWh to Achieve 10.0% Year 1 ROE

Gas CCGT Nuclear Wind

Market Price per MWh 81$ 56$ 60$

Cost per kW 900 5,116 1,900Capital 900 5,116 1,900% Equity 20% 20% 20%Assumed Debt 720 4,093 1,520Assumed Equity 180 1,023 380

Assumed Depreciable Plant Life 30 36 25Interest rate % 8.0% 4.5% 8.0%Tax Rate % 38% 38% 38%

Megawatt (000's) 1.000 1.000 1.000Assumed Capacity Factor 70% 90% 35%Megawatt-hours 6.132 7.884 3.066Fuel cost per MWh 56.00$ 5.00$ -$Heat Rate 7.000 10.000 0.000$ Cost per MMBtu 8.00$ 0.50$ -$$ Cost per Mcf 8.00$Heat Content 1,000,000

Margin 153 405 183O&M Per Kw-year 30$ 79$ 14$Operation & Maintenance Costs (30) (79) (14)Depreciation & Amortization (30) (140) (76)Property Taxes (7) (38) (14)Operating Income 87 148 79

Interest Expense (58) (184) (122)Pre-tax Income 29 (36) (43)Income Tax (11) 14 16PTCs 125 64Net Income 18 102 38Cash Flow 48 243 114

ROE 10.0% 10.0% 10.0%Source: Jefferies & Company, Inc. estimates

EXHIBIT 4: SCENARIO ANALYSIS OF ALTERNATIVE ENERGY SOURCES



Page 185 of 212

Scenario 2 : No PTCs or Nuclear Government Loan Guarantees

Gas CCGT Nuclear Wind

Market Price 82$ 56$ 60$

Cost per kW 900 5,116 1,900Capital 900 5,116 1,900% Equity 20% 20% 20%Assumed Debt 720 4,093 1,520Assumed Equity 180 1,023 380

Assumed Depreciable Plant Life 30 36 25Interest rate % 8.0% 8.0% 8.0%Tax Rate % 38% 38% 38%

Megawatt (000's) 1.000 1.000 1.000Assumed Capacity Factor 70% 90% 35%Megawatt-hours 6.132 7.884 3.066Fuel cost per MWh 56.00$ 5.00$ -$Heat Rate 7.000 10.000 0.000$ Cost per MMBtu 8.00$ 0.50$ -$$ Cost per Mcf 8.00$Heat Content 1,000,000

Margin 158 405 183O&M Per Kw-year 30$ 79$ 14$Operation & Maintenance Costs (30) (79) (14)Depreciation & Amortization (30) (130) (76)Property Taxes (7) (38) (14)

Operating Income 87 158 79

Interest Expense (58) (327) (122)Pre-tax Income 29 (169) (43)Income Tax (11) 64 16PTCs 0 0Net Income 18 (105) (27)Cash Flow 48 25 49

ROE 10.0% -10.3% -7.0%


EXHIBIT 5: SCENARIO ANALYSIS OF ALTERNATIVE ENERGY SOURCES ASSUMING NO SUBSIDIES



Page 186 of 212

Enrichment

The cost of nuclear fuel is often an after-thought in assessments of nuclear plant economics, representing roughly 0.5 cents/kWh. We estimate, however, that process economics that secure a 10% return for investments in new enrichment capacity, without any government subsidy, could add as much as 0.15 cents/kWh fuel costs, comparable to a $50/lb swing in the price of uranium. To put this in context, we estimate that a 0.25 cents/kWh change in fuel costs has a 25bps impact on our projection of nuclear power plant IRRs: the equivalent of a $50/kW change in EPC cost or a 4% change in our interest rate assumption.

Contract Uranium SWU $/lb $100 $125 $150 $175 $200 $225 $40 0.360 0.418 0.476 0.534 0.592 0.649 $50 0.388 0.446 0.504 0.562 0.62 0.677 $60 0.417 0.475 0.533 0.591 0.649 0.706 $70 0.445 0.503 0.561 0.619 0.677 0.734 $80 0.473 0.531 0.589 0.647 0.705 0.762

$100 0.543 0.601 0.659 0.717 0.775 0.832 $150 0.614 0.672 0.730 0.788 0.846 0.903


To illustrate this, it is worth revisiting the nuclear fuel cycle. Utilities contract for mining companies to supply the ore and processors to convert and enrich the ore until it is suitable to use as nuclear fuel. Enrichment activity is measured in SWU, which is simply a measure of the amount of effort needed to enrich uranium for a utility. In other words, the more efficiently uranium is processed, the more SWU is required and the less uranium ore is needed to make the same amount of fuel.

Supplying a typical 1GW light water reactor encompasses the following steps:

• Mining 20,000 t of 1% uranium ore.

• Milling this ore to produce 230 t of uranium oxide (195 t U, with 0.71% U-235)

• Converting this into 288 t of uranium hexafluoride (195 t U)

• Enriching this to create 35 t of uranium hexafluoride (24 t enriched U, with 4% U-235) and 253 t of waste ‘tails’ (0.25% U-235). This step can require 100,000-120,000 SWU.

• Fabricating 27 t of uranium dioxide (24 t enriched U)

• Removal of 27 t of used fuel, including 23 t uranium, 240kg of plutonium and other fission products.

Another way to look at the process economics is to start with the cost of making a single kg of uranium fuel suitable for use in a reactor. We estimate that enrichment services currently represent roughly 64% of the total cost of nuclear fuel.

Another factor, which we assume is static for the sake of this analysis, is the cost of decommissioning and fuel disposal, which we estimate at $0.02-$0.03/kWh—significantly more on a $/kWh basis than the cost of the fuel itself.

EXHIBIT 6: SENSITIVITY OF IMPLIED FUEL PRICE (CENTS/KWH) TO COST OF SWU AND URANIUM



Page 187 of 212

Process Step Units

required Cost/Unit

Total Cost Comment

Obtain U308 (kg) 8.9 $39 $346 19% of total cost. Assume long-term contract costs

1/3-1/5 of spot price. Conversion (kg of U) 7.5 $10 $75

Enrichment (SWU) 7.3 $159 $1,161

64% of total cost. Boosting enrichment levels (i.e. using more SWU) helps reduce the near-term

pressure of higher uranium prices. Fuel fabrication ($/kg of fuel) $240

Total 1kg of UO2 $1,822

Yields 315,000-360,000 kWh, or $0.005-$0.006/kWh

Other Costs

% of Initial

Capital Cost $/kWH Comment

Decommissioning 9%-15% $0.01-$0.02

Typically less than 5% of the cost of electricity produced once discounted.

Spent fuel storage/disposal $0.01 $0.01 levy in U.S. Can be as much as 10% of total cost/kWh


For the sake of this scenario analysis, four points about the enrichment market are relevant.

0.20% 0.25% 0.30% 0.35% 0.40%

Aug-06

Sep-05Apr-05

Jan-05Dec-03

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

Optimal Trails

Rat

io:$

/KgU

asU

F6to

$/SW

U

Oct-00

0.20% 0.25% 0.30% 0.35% 0.40%

Aug-06

Sep-05Apr-05

Jan-05Dec-03

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

Optimal Trails

Rat

io:$

/KgU

asU

F6to

$/SW

U

Oct-00

Source: WNA

First, nuclear enrichment acts as a safety valve for the nuclear industry. Providers of nuclear enrichment such as USEC (USU, $3.44, Buy) can change the efficiency with which they process uranium for the utilities. For many years, the industry’s long-term contracts allowed the utilities to arbitrage between the cost of the raw uranium ore and SWU prices. In other words, to reduce uranium needs, utilities can use their uranium more efficiently (which requires more SWU).

Second, we believe the underlying dynamics have changed and the safety valve is broken. For decades the supply/demand balance favored utilities, as the cost of enriching their fuel was modest. By the middle of the next decade, however, the enrichment industry could be effectively sold out. The pace with which the enrichment industry has announced capacity expansions has not kept up with the pace of new reactor announcements.

EXHIBIT 7: KEY STEPS IN NUCLEAR FEEDSTOCK COSTS ($/KG OR $/SWU)

EXHIBIT 8: RATIO BETWEEN URANIUM AND SWU PRICES AND OPTIMAL TAILS ASSAYS (I.E., AMOUNT OF WASTED URANIUM): THE EXAMPLE FROM 2000-2006



Page 188 of 212

2002 2006 2015 Areva 10.8 10.8 8.5 Urenco 5.9 9 15.5 USEC 8 11.3 3.5 JNFL 0.9 1.05 1.5 Tenex 20 25 33 China 1 1 1 Other 0 0.3 0.3 Total 46.6 58.45 63.3 Demand 35 48.43 57-64

Source: WNA, Jefferies & Company, Inc. estimates

Third, as a consequence of the first two points, enrichment costs have more than doubled since 2000. Contract terms have tightened. Providers of enrichment services have introduced indexed escalation to offset volatility in energy and other input costs. Also, the industry is shifting away from large volume gaseous diffusion plants to modular centrifuge technologies. This lets enrichment companies set contract terms at specific tails assays, which curtails utilities ability to offset rising uranium prices. This should become more evident as the new contracts come into effect in the 2009-2015 period.

$60

$80

$100

$120

$140

$160

Jan-

95

Jan-

96

Jan-

97

Jan-

98

Jan-

99

Jan-

00

Jan-

01

Jan-

02

Jan-

03

Jan-

04

Jan-

05

Jan-

06

Jan-

07

Jan-

08

SWU Price

Source: Tradetech

Fourth, the enrichment industry needs to recoup an adequate return on investment on the new centrifuge capacity that is currently under construction.

We have constructed a greenfield model for a new enrichment facility using USEC’s centrifuge technology—but excluding $525m in R&D costs that USEC has invested to develop the centrifuge technology in recent years. We have also disregarded the $3bn or more that the U.S. government has already invested in developing centrifuge technology over the past couple of decades.

Without a government subsidy, and assuming 20% equity financing, we estimate a greenfield SWU facility would generate a 7% ROIC in its fifth year and a peak ROIC of 9%. Alternately, a stand-alone facility would generate an unlevered IRR of 4.4% and a levered IRR of 5.4%.

Such a project, however, has little bearing on real world economics. For example, securing a license for a new greenfield facility is extremely difficult. Regulatory hurdles represent a significant barrier to entry for new participants, and opposition by consumers and environmentalists can significantly constrain site selection.

A more practical comparison would be to compare USEC’s first tranche of centrifuge capacity at its $3.5bn, 3.8m SWU, American Centrifuge project with the economics on a capacity expansion at the same site. For the first tranche, we have included the $525m of R&D invested in developing the centrifuge technology over the past few years. We have also factored in $200m invested in plant infrastructure. Assuming the current SWU price of $159, these assumptions lower the estimated long-term ROIC to 5%.

EXHIBIT 9: WORLD ENRICHMENT SUPPLY/DEMAND (M SWU/YEAR)

EXHIBIT 10: SWU PRICES, 1995-2008 ($/SWU)



Page 189 of 212

Third, when we compare a stand alone facility using USEC’s centrifuge technology, with 3.8m SWU capacity costing $3.5bn to build, with a facility using USEC’s technology where a second tranche of capacity has been added, the peak ROIC for the facility improves to 11%. Fixed cost absorption, opportunities to leverage existing staff, and process improvements and better centrifuge design would likely be the main levers driving such a difference in return profiles.

In each case, however, the implication is that USEC either needs to content itself with suppressed ROIC, or SWU prices should be flat-to-higher over the next decade.

There are three scenarios to achieve higher returns on capital on these projects:

• SWU prices could increase. We estimate fuel prices could rise as much as 32%, or 0.16 cents/kWh, in order to incentivize USEC to proceed with its second tranche of capacity. We estimate that each $500m cost overrun on building the first tranche of capacity would increase the required SWU price by $10 and fuel prices by 0.02 cents/kWh.

Unlevered SWU price needed

for 10% ROIC in Year 10

Implied Fuel Price

Year 5 Year 10 Year 15 IRR ($/SWU) ($/kwh) Current $159 $0.0050 Theoretical 6.6% 8.3% 9.0% 4.4% $171 $0.0052 USEC tranche 5.0% 5.0% 5.0% 3.1% $235 $0.0066 With expansion 5.0% 12.0% 11.4% 8.3% $145 $0.0047


• The government could provide a direct subsidy. We estimate the first USEC tranche would require a $1.85bn subsidy to achieve a 10% ROIC, whereas with a second facility expansion USEC could achieve a 10% ROIC while delivering a net reduction in fuel prices to its customers even without a government subsidy.

• USEC has applied for part of a $2bn government loan guaranty program authorized by the 2007 Energy Act. Winning some portion of this program could significantly reduce its cost of debt. While this would not improve returns on capital, it would improve the IRR on the project and the estimated terminal ROA.

Loan Guaranty 0% 25% 50% 100% ROIC 5.0% 5.0% 5.0% 5.0% ROA 4.2% 4.2% 4.2% 4.2% ROE 10.5% 10.1% 9.5% 9.2%

Unlevered IRR 3.1% 3.1% 3.1% 3.1% Levered IRR 0.3% 0.6% 1.2% 1.5%


EXHIBIT 11: COMPARATIVE ROIC AND IRR ANALYSIS FOR NEW CENTRIFUGE CAPACITY: THEORETICAL GREENFIELD VS. FIRST AND SECOND USEC TRANCHES

EXHIBIT 12: RETURNS FROM FIRST TRANCHE OF USEC’S CENTRIFUGE ENRICHMENT CAPACITY WITH VARIOUS LOAN GUARANTIES (% OF TOTAL PROJECT COST)



Page 190 of 212

EventThis primer provides an overview of project finance for renewableenergy investors, with a focus on the pros and cons, as well as asurvey of key concepts and requirements, including tax incentivesand monetization strategies in the renewable energy sector, andother key structuring considerations in determining whether to projectfinance.

Key Points• Project finance has emerged as a leading way to finance large

infrastructure projects that might otherwise be too expensive orspeculative to be carried on a corporate balance sheet.

• The basic premise of project finance is that lenders loan money forthe development of a project solely based on the specific project'srisks and future cash flows. As such, project finance is a method offinancing in which the lenders to a project have either no recourseor only limited recourse to the parent company that develops or"sponsors" the project.

• For equity investors, the appeal of project finance is that it canmaximize equity returns, move significant liabilities off balancesheet, protect key assets, and monetize tax financing opportunities.A wide range of commercial and legal issues must be addressed tosecure adequate returns. Tight credit markets exacerbatecompetition for long-term financing, so even small differences indeals can impact the availability of financing or reduce leverage.

• Project financing became particularly important to projectdevelopment in emerging markets, with participants often relying onguarantees, long-term off-take or purchase agreements, or othercontractual relationships with the host sovereign or its commercialappendages to ensure the long-term viability of individual projects.These were typically backstopped by multilateral lending agenciesthat mitigated some of the "political" risks to which the projectlenders were exposed. Analogies to alternative energy projects helpinvestors de-risk higher-risk new technologies.

October 7, 2008


Clean TechnologyProject Finance Primer

Investment SummaryClean technology investments often combine capital intensity withnew technologies. Securing project finance can prove to be a criticalstep for commercialization. Project finance works best when you havelong-term off-take agreements with quality-credit counterparties (e.g.,power purchase agreements), but commodity-based projects that sellinto open markets (e.g., biofuels) can also benefit.


Chris Groobey, Baker & McKenzie LLP(202) 835-4240, [email protected]

Nathan Read, Baker & McKenzie LLP(202) 835-1668, [email protected]

For further details, please contactChris Groobey or Nathan Read.

This report was prepared by the RenewableEnergy Group of Baker & McKenzie LLP*.

*In addition to Chris Groobey and Nathan Read,the following attorneys at Baker & McKenziecontributed to the preparation of this report:Marc Paul, Maria Sendra and Kenji Funahashi (equity),Susan Stone and Michael Snider (tax) andSteve Otillar (project finance).


EXHIBIT 1: SIMPLIFIED LOOK AT AFTER-TAX RETURNS FROM PROJECT FINANCE: “INVESTOR A” CONTRIBUTES PROJECT EXPERTISE, “INVESTOR B” TECHNOLOGY, AND “CREDITOR C” 70% OF TOTAL CAPITAL

Investor A Investor A Benefits:50% stake -- Attractive IRR despite lack of IP in new technology

Investor B Benefits:Project -- No recourse on Project$100 Invested Capital -- 50% equity control for 10% of capital10% IRR -- 33% IRR vs. 22% on 70/30 debt finance

Investor B50% stake

Creditor C

16% IRR

$20 cash, Project expertise

$70 7.5% cost of debt

$10, Technology

33% IRR

Source: Baker & McKenzie LLP, Jefferies & Company, Inc. estimates

EXHIBIT 2: COMPARISON OF “GOOD DEAL” VS. “KILL THE DEAL” CHARACTERISTICS WHEN CONSIDERING PROJECT FINANCE

“Good Deal” “Kill the Deal” Lender perspective 1. Size of project >$50m <$25m 2. EBIT/Interest > 3x < 1x 3. Contracted revenue Yes None 4. Liquidation value Covers debt None 5. Technology risk None First proof of concept 6. Contracted suppliers Blue chip No contracts 7. Control of uses of cash flow Creditor Sponsor 8. Quick exit for Sponsor No Yes 9. Oversight Creditor veto None

Equity perspective 1. Risk of a "cram down" None 100%2. Strategic partners In place Uninterested 3. Path to profitability Certain Doubtful 4. Control Secured None 5. Conflicts of interest None Yes 6. Anti-dilution rights, etc. Yes No 7. Commercialization timeline Matches exit strategy None 8. Liquidity event Near-term Never 9. Returns on capital 500bps > WACC < WACC

Source: Baker & McKenzie LLP

Industrial Biotech



Page 192 of 212

Project Finance Primer

This primer undertakes an overview of project finance for the renewable energy investor or developer, with a focus on the pros and cons of project finance, as well as an examination of some of its more fundamental characteristics and requirements, including how to raise equity capital. Tax incentives and monetization strategies in the renewable energy sector, and key structuring considerations in determining whether to project finance, are also examined.

Part I of the primer introduces project finance to those that may be less familiar with the concept, and asks questions that will assist investors and developers in determining whether project finance is appropriate for their renewable energy projects. Part II sets out the legal and contractual structure that will facilitate project financing. Part III describes the process of obtaining equity investment and some of the important options and considerations that companies may have in that process. Part IV provides a more in-depth look at what a typical renewable energy project financing looks like, including fundamental structural components that characterize any project finance transaction. Finally, Part V outlines key tax incentives currently available in the renewable energy industry, as well as monetization strategies that may be useful for earlier-stage energy companies unable to directly utilize such tax incentives.

Given the breadth of the current renewable energy landscape, this primer focuses on a hypothetical solar generation facility (“Solar Project”) as the primary case study with discussions of other renewable energy projects (wind power and biofuel projects in particular) as appropriate. Typically, Solar Projects are either: (i) concentrating thermal solar generation facilities, which use solar power to heat liquid or another substance to power a turbine or engine (“CSP”), or (ii) photovoltaic solar facilities that harness solar power for conversion directly into electricity (“PV”). CSP facilities are limited in geographic scope to sunny, arid climates at relatively low latitudes (for example, the southwest United States) but are larger in scale and usually connected to the electric grid. PV facilities, on the other hand, can distribute electricity directly to a host while bypassing the electric grid and have historically been quite small in size (under 2 MWs), although the size of PV facilities is increasing, particularly in Europe and in the United States as well. While not a focus of this primer, it should be noted that many states have implemented renewable portfolio standards that mandate the use of renewable power (even specifically solar power in some cases) in a certain percentage of total consumption by a given date, and other incentive programs that encourage the use of renewable energy generally and solar power in particular.

The discussion of the key tax incentives in Part V not only covers the principal incentives available to a Solar Project, but also addresses tax incentives available to other renewable sources of energy. Finally, it should be noted that once the contracts related to a project are negotiated (which is described in Part II), the mechanical aspects of raising equity and project financing are likely to be similar across various renewable technologies, although investor enthusiasm and financing prices and terms are likely to vary significantly across technologies at any given time.

I. Introduction to Project Finance

A. What Is Project Finance?

The basic premise of project finance is that lenders loan money for the development of a project solely based on the specific project’s risks and future cash flows. As such, project finance is a method of financing in which the lenders to a project have either no recourse or only limited recourse to the parent company that develops or “sponsors” the project (the “Sponsor”). Non-recourse refers to the lenders’ inability to access the capital or assets of the Sponsor to repay the debt incurred by the special purpose entity that owns the project (the “Project Company”). In cases where project financings are limited recourse as opposed to truly non-recourse, the Sponsor’s capital may be at risk only for specific purposes and in specific (limited) amounts set forth in the project financing documentation.

Project financing has been used in various ways for many years, but in the 1970s and 1980s it emerged as a leading way of financing large infrastructure projects that might otherwise be too expensive or speculative for any one individual investor to carry on its corporate balance sheet. Project financing has been particularly important to project development in emerging markets, with participants often relying on guarantees, long-term off-take or purchase agreements, or other contractual relationships with the host sovereign or its commercial appendages to ensure the long-term viability of individual projects. These were typically backstopped by multilateral lending agencies that mitigated some of the “political” risks to which the project lenders (and, sometimes, equity investors) were exposed.

Industrial Biotech



Page 193 of 212

B. What Underpins Project Finance?

As a general (if not universal) rule, lenders will not forgo recourse to a project’s Sponsor unless there is a projected revenue stream from the project that can be secured for purposes of ensuring repayment of the loans. In the case of large wind and solar power projects, this revenue is typically generated from a power purchase agreement (“PPA”) with the local utility, under which the project may be able to utilize the creditworthiness of the utility to reduce its borrowing costs. While the wind power market has matured significantly in the past five years, leading to the successful project financing of “merchant” projects in the absence of long-term PPAs, Solar Projects are generally not yet able to be project financed in such a manner. In merchant power projects, the lenders are able to receive assurance of the project’s ability to repay its debt by focusing on commodity hedging, collateral values, and the income to be produced based on historical and forward-looking power price curves and fully developed markets. In non-power generation contexts, the project’s revenue stream may be a long-term operating agreement (e.g., in the case of toll roads), a capacity purchase agreement (e.g., in the case of transmission lines), a production sharing agreement (e.g., in the case of oil field development), or a series of short-term and spot sales into commodity markets (e.g., in the case of biofuels projects).

While project finance lenders clearly prefer a long-term contract that ensures a relatively consistent and guaranteed revenue stream (including assured margins over the cost of inputs), in the context of some industries, lenders have determined that sufficient revenues to support the project’s debt are of a high enough probability that they will provide debt financing without a long-term off-take agreement. Examples of such industries are power generation facilities in more mature deregulated markets (where operating histories of similar plants demonstrate cost recovery) and casinos in Macau (due to the sheer enormity of revenue projections). Solar Projects, due to their peak period production, high marginal costs, and lack of demonstrated merchant capabilities, are not at this time viewed as “project financeable” without PPAs that cover all or substantially all of their output. Solar Projects’ lack of merchant viability is exacerbated by the fact that the southwest United States (the region most appropriate for utility-scale solar power development) does not have a mature merchant power market that functions in the absence of long-term bilateral sales agreements. The dependence of large-scale solar projects on the PPA model is not expected to change in the short to intermediate term.

C. When to Project Finance?

One of the primary benefits of project financing is that the debt is held at the level of the Project Company and not on the corporate books of the Sponsor. When modeling projects and projected income, the internal rate of return of Sponsors and other project-level equity investors can increase dramatically once a project is fully leveraged. Sponsors are frequently able to recover development costs at the closing of the project financing and put their money into other projects. Another benefit of project financing is the protection of key Sponsor assets, such as intellectual property, key personnel, and investments in other projects and other assets, in the case of the Project Company’s bankruptcy, debt default, or foreclosure. Moreover, project financing allows for a wide variety of tax structuring opportunities, particularly in the context of monetizing tax incentives (discussed further in Part V). On the other hand, project financing is document-intensive, time-consuming, and expensive to consummate. It is not atypical that administrative and closing costs, when factoring in lenders, consultants, and attorneys fees for all parties, equal several percentage points of the amount of the loan commitment. Moreover, project financing imposes significant operating restrictions on each Project Company, including its ability to make equity distributions to the Sponsor prior to the payment of operating expenses, debt service, and a percentage “sweep” of additional cash flow (discussed further in Part IV). The result is that the decision of whether to reinvest cash flow in the project does not rest solely with the Sponsor.

Given the pros and cons of project finance, the most relevant initial inquiry for an investor or developer may be when is project financing possible or most appropriate? The following questions should be useful in determining if project financing is a realistic opportunity for any given company:

- Is there an individual project or group of projects of a sufficient size to make either a standalone or portfolio project financing worthwhile? Typically lenders will be reluctant to provide project financing if the total amount of debt is less than US$50 million and, preferably, US$100 million.

- Will there be a revenue stream from the project large enough to support a highly leveraged debt financing? This is a prerequisite for project financing.

- Will the receipt of revenue be enforceable under contractual rights against a creditworthy party? This is not necessarily a prerequisite for all project financings, but the absence of a contract, or questionable creditworthiness of the purchaser, will prompt lender skepticism and necessitate thorough due diligence regarding future revenue projections.

- Will there be physical assets sufficient to ensure lender repayment in case of foreclosure? Lenders will want to know that even if the Project Company’s projected revenue stream does not materialize,

Industrial Biotech



Page 194 of 212

they will be able to foreclose on the project’s assets sufficient in value to “make themselves whole,” either by selling the project outright or operating it until the debt is repaid.

- Is there a significant level of technology risk? While in many project financings, technology may be relatively new or cutting edge, project finance lenders almost never want to be the first to finance an untested technology. Demonstrated successful use in some context will often be necessary to secure project financing.

- Does the project have contractual relationships with reputable companies for services key to the success of the project or the technology it employs? Lenders will be less likely to lend to a project the success of which depends solely on a few talented individuals who may depart, leaving the project unable to meet its potential.

- Is the Sponsor ultimately willing to “risk the project”? In other words, once project financing is completed, the Sponsor loses the ability to determine how the vast majority of the project’s revenue is spent. In the event a project becomes uneconomic and unable to service its debt, the only option besides refinancing the debt may be to turn over the project to the lenders (voluntarily or involuntarily), with the corresponding loss of the Sponsor’s investment in the project.

- Is the Sponsor looking for a quick exit? Once project financed, divestiture opportunities are complicated by the requirement of lender consent, and potential purchasers will be thoroughly examined by lenders for development and operational expertise as well as creditworthiness.

- Are Sponsors willing to grant rights of high-level oversight regarding the project’s development and operation to project finance lenders? In many cases the interests of the Sponsor and the lenders will be aligned, and lenders will tend to defer to the Sponsor’s developmental expertise. On the other hand, lenders must be viewed as additional project partners, with veto rights over many significant decisions.

Assuming project financing is a viable option, Part II provides a roadmap to structuring a project financing transaction.

Industrial Biotech



Page 195 of 212

II. Establishing a Project Structure and Negotiating Project Agreements

A. Project Structure

The project finance structure revolves around the creation of the Project Company that holds all of the project’s assets, including all of its contractual rights and obligations. The Project Company is usually a single-member limited liability company, although in some cases it may be a limited partnership.

In most cases, the equity interest in the Project Company will be held by at least one intermediate holding company, usually a limited liability company (the “Holdco”), created for the purpose of pledging the Project Company’s equity to the lenders in the eventual project financing. While the Holdco will have a separate legal identity and officers, typically it will not have any business apart from holding the equity of the Project Company. This structure allows for most liability to be contained at the bankruptcy-remote Project Company level, and thus insulates the Sponsor (including equity investors in the Sponsor) and the Holdco from liability to either the Project Company’s contractual counterparties (“Counterparties”) or to its lenders. In order to ensure that the Project Company is treated as a separate legal entity, it will be necessary to have governance mechanisms at the Project Company level that are independent, including designated officers, at least one independent director, and internal controls and procedures designed to preserve a legal entity distinct from the Sponsor and the Holdco.

B. Project Agreements Overview

As a general matter, all contracts related to the development, construction, ownership, and operation of the project will be entered into by the Project Company (“Project Agreements”). If development-stage contracts have been executed by the Sponsor or one of its affiliates, it is important that those contracts allow for their assignment to the Project Company once the Project Company has been established for the purposes of pursuing project financing.

In addition to the external Project Agreements, there may be several intercompany agreements between the Project Company and the Sponsor or its affiliates. These may include an Operation and Maintenance Agreement (“OMA”), an Administrative Services Agreement (“ASA”), and a Technology License Agreement (“TLA”), often with affiliates of the Sponsor created specifically for the purpose of providing administrative support, operation, and maintenance services and holding the intellectual property for the benefit of one or more of the Sponsor’s projects. In other cases, unrelated third parties may provide these services to the Project Company. If intercompany agreements are used, they should be structured in such a manner as to roughly track the commercial terms that the Sponsor could obtain with an unrelated third party providing the same services.

Intercompany agreements can also have a significant impact on the total return of a project to its investors, so their economic terms must be carefully crafted. Assuming the OMA, ASA, and TLA are entered into with Sponsor affiliates, they permit the affiliates to extract “arms length” fees for the provision of key services and technology to the Project Company on a monthly or quarterly basis, sometimes even prior to repayment of debt. The intercompany-agreement structure also allows the Sponsor, if the project fails following the project financing, to retain all of its employees that provide services to the Project Company, thereby ensuring that key employees (and know-how) will not be lost to lenders or a subsequent purchaser out of foreclosure. In such a scenario, the TLA will also allow the Sponsor to retain ownership of its technology subject only to a license right on the part of the Project Company which may no longer be affiliated with the Sponsor. These are especially critical points where the Sponsor has multiple projects that may utilize the same technology, support equipment, and personnel. In addition, the OMA, ASA, and TLA provide the Project Company's lenders contractual certainty (through the agreements themselves as well as the corresponding consents to collateral assignment (discussed further in Part IV below) that key services will continue if the Project Company defaults, thereby increasing the likelihood of the efficient development, construction, and operation of the project and the preservation of the value of the lenders' collateral.

There are many other Project Agreements that are typically executed during the course of developing and constructing a renewable energy project. The Project Agreements may include one or more PPAs, which may have an income stream payable from an off-taker for energy payments, capacity payments, or both; an Engineering, Procurement, and Construction Agreement (“EPC Agreement”); or, as is the more recent trend with contractors unwilling to provide lump sum turnkey prices, EPC management contracts (“EPCM Agreements”), various agreements for the purchase of equipment necessary to successfully construct and operate the project (although in some cases these will be assumed by the EPC contractor in what is referred to as a “full wrap” EPC Agreement). These include, for example, for solar panels, steel, and steam or wind turbine generators, a Site Lease Agreement (if the project’s land is not owned by the Project Company itself); a Renewable Energy Credit Agreement (in states where applicable); an Interconnection Agreement (for projects tied to the electricity grid); agreements for the provision of utility services; agreements for the provision of feedstock commodities (in the case of biofuels); agreements including equity flip structures to take

Industrial Biotech


advantage of federal tax incentives discussed in Section V below; and other Project Agreements necessary or desirable to develop, construct, own, or operate the project. In some cases certain byproducts of production may be sold in addition to the primary product (for example, steam as a byproduct of co-generation power projects or high protein distillers grain as a byproduct of ethanol production).


Page 196 of 212

EXHIBIT 3: TYPICAL PROJECT FINANCE STRUCTURE


C. Key Project Agreement Terms

In the process of negotiating the Project Agreements it will be necessary to consider key project finance principles to prevent having to revisit contractual terms at the lenders’ behest in the course of financing the project. One overriding concept is that lenders will own (and likely seek to immediately transfer) the Project Company in the case of foreclosure, thus will insist on contractual rights and terms that ensure a seamless transition to the lender or subsequent owner. To this end, the project lenders will require consents to collateral assignment (“Consents”) for their benefit with some if not all of the Counterparties. Therefore, provisions that prevent assignment without Counterparty consent should be omitted from Project Agreements. Inclusion of contractual language that obligates the Counterparty to cooperate with the Project Company and its lenders in the course of the financing process will not only expedite the process of negotiating the Consents, but also will reduce the scope for Counterparty intransigence in the context of the project financing.

The commercial terms of the PPA and the EPC Agreement, together with the market and technology risks, will largely determine whether lenders view the project as “financeable.” Foremost among considerations related to the PPA will be whether or not there is a guaranteed revenue stream (usually energy payments from the actual production of power) from a creditworthy purchaser that will be sufficient to support the economics of the project, thereby ensuring prompt repayment of debt and mitigating the risk of default. The PPA term should also be sufficient in length to fully amortize the contemplated project debt. In contrast to most smaller distributed generation projects, where an off-taker pays for only the power that is produced, utility scale solar generation facilities may have “take or pay” PPAs where the utility is still required to pay the Project Company even if a certain level of power is not purchased. In the case of Solar

Equity Investors

Sponsor Project-Level Equity Investors Lenders

Project Companyaka “The Borrower”

Inter-connecting

Utility

Lenders

PurchasingUtility/

Offtaker

Lenders

EquipmentSuppliers

Lenders

Contractor

Lenders

RECPurchaser

Lenders

O&MServiceProvider

Lenders

AdminServicesProvider

Lenders

TechnologyProvider

Lenders

EquityContribution $

$

Operating or ShareholdersAgreement


Operating or LimitedPartnershipAgreement

EquityContribution $ $


Loans $ $LoanDocumentation

TechnologyLicenseAgreement

AdminServicesAgreement

O&MAgreement

RECPurchaseAgreement

SupplyContracts

EPC orEPCMAgreement

PPAInter-connectionAgreement

Consent Consent Consent Consent Consent Consent Consent Consent

$

Equity Investors

Sponsor Project-Level Equity Investors Lenders

Project Companyaka “The Borrower”

Inter-connecting

Utility

Lenders

PurchasingUtility/

Offtaker

Lenders

EquipmentSuppliers

Lenders

Contractor

Lenders

RECPurchaser

Lenders

O&MServiceProvider

Lenders

AdminServicesProvider

Lenders

TechnologyProvider

Lenders


$

Operating or ShareholdersAgreement



EquityContribution $ $


Loans $ $LoanDocumentation

TechnologyLicenseAgreement

AdminServicesAgreement

O&MAgreement

RECPurchaseAgreement

SupplyContracts

EPC orEPCMAgreement

PPAInter-connectionAgreement

Consent Consent Consent Consent Consent Consent Consent Consent

$

Typical Project Finance Structure

Industrial Biotech


Projects, utilities are less likely to deliver capacity payments because power is generally produced only during daylight hours. However, distributed generation projects may benefit from more certain payments because the distributed project produces all of the power for a given host.

If a project does not have a PPA or other off-take contract, demonstrated merchant operating histories of similarly situated plants (more relevant in the context of wind projects) will be necessary to convince lenders of the reliability of forecast ratios. Even with long-term PPAs, lenders will still look for additional data to support viability such as meteorological wind data for wind power sites over the course of one to two years, often at installed hub-heights, or long-term temperature and sun data for Solar Projects. The trend in biofuels project financings is also moving toward


Page 197 of 212

contracted off-take arrangements with a creditworthy purchaser for all of a plant’s production, although historically plants that have only marketing agreements and no fixed off-taker have also been project financed. We expect that, at least in the short and intermediate terms, most project-financed Solar Projects will have PPAs for a significant portion, if not all, of their generated power. While PPAs for large-scale CSP projects will generally be far more complex than those for smaller distributed PV projects, in either case, the core economic terms will determine the lenders’ view of whether a project or a portfolio of projects is viable from a revenue perspective and, accordingly, “financeable” on favorable terms.

To the extent a project is not fully constructed by the time project financing is sought, EPC Agreements will be an integral part of the financing analysis and pricing. While larger developers may be able to finance an entire project on balance sheet, and subsequently refinance the development to free up invested capital, most developers use project finance to construct and operate their projects. Where construction risk is present, lenders will generally seek corporate parent guarantees and performance bonds that ensure the performance of the contractor is as close to budget and schedule as possible. Warranties of appropriate substance and duration as well as subsequent maintenance coverage regarding the EPC work and the equipment purchased will be necessary to convince the lenders that significant unbudgeted expenses will not be incurred by the Project Company. This is particularly the case with newer and untested technology even if operationally superior to previous generation technology. Liquidated damage coverage (pre-agreed payments made by the contractor) for schedule and performance delays, inefficiency, or equipment failures also reassure lenders that a project has the necessary protection against delays or performance defects that are within the EPC contractor’s control. How much of the risk an EPC contractor accepts for cost overruns and design or installation defects, when viewed with other contractual terms, will affect the lenders’ view of whether a project is “financeable” and at what cost. For example, a project that is not financeable at 80% debt due to certain off-take or technology risks may be financeable with 40%–60% equity because the lenders are taking less risk with a higher level of capital pre-paid into the project. Dedicating sufficient resources at the negotiation stage of a PPA or EPC Agreement to achieve commercial and contractual terms as favorable as possible will usually pay dividends at the financing stage by saving not only money, but also costly renegotiation and valuable time toward project completion. It is critical that professional consultants be involved not just in final documentation, but at or prior to the term sheet stage in order to extract the most value for the Sponsor.

III. Raising Equity

Venture capital (early stage funds) and private equity (later-stage funds) investors also serve as attractive sources for Solar Project capital raising, as an increasing number of funds are investing in solar energy. The large sums of capital required to initiate and complete Solar Projects drive not only the selection of appropriate equity investors, but also the structure of such investments. Therefore, Sponsors and investors evaluating equity solar investments should consider the following action items:

A. Conduct an internal assessment of capital and budgeting strategies for the investment.

Sponsors and investors should conduct an internal evaluation of whether an equity investment would best serve their respective strategic objectives. Salient considerations include:

- Is the company’s technology reliable enough to be considered financeable, and is there a realistic potential pool of equity investors from which to draw?

- Are there any gaps in the Sponsor’s existing organizational structure and operations that an equity investor would want filled before engaging in substantive discussions and/or closing an equity round of financing? For example, given the complexity involved in successfully executing Solar Projects, equity investors will look for experienced management with skill and connections within the industry, as well as

Industrial Biotech


potentially requiring contractual commitments from relevant third parties in the supply chain and customer base.

- How much capital investment is realistically required for the Solar Project? Has Sponsor management, on the one hand, conducted a thorough analysis of the timing and amount of future capital needs and relevant burn rates, and has the investor and its syndicate, on the other hand, assessed whether its proposed financing will be sufficient to either execute the Solar Project or bridge the Sponsor towards its next round of investment?

- What type of investment is ideally suited for the particular Solar Project — e.g., is the Sponsor seeking passive investment, or an active strategic partner that will add value to the organization (as discussed in greater detail below)?

- How would an equity investment impact the Solar Project’s existing grants, tax treatment, eligibility for applicable federal and state incentive programs, and contractual obligations?


Page 198 of 212

- As equity investors become shareholders, and in many cases, directors of the Sponsor, how much control is the Sponsor willing to give to the investor, and what level of control does the investor desire in order to have an active voice within the organization?

B. Determine whether the investment will add value to the Solar Project.

Unless a Sponsor is seeking a purely passive equity investment, the Sponsor and its investor should conduct a thorough assessment of the investor’s role in driving value to the enterprise by, among other things:

- Reviewing the investor’s existing portfolio companies to determine whether the investor has previously invested in Solar Projects or has other relevant experience with alternative energy investments. It is important to find investors who understand the longer time period required to execute and obtain a return on investment from Solar Projects;

- Meeting with the investor’s key decision makers to assess how the investor will add value in addition to the capital infusion — e.g., through participation on the board of directors, introductions to potential customers, assistance in financial forecasting and planning, and guidance in analyzing potential liquidity events; and

- Assessing potential conflicts of interest that may arise to the extent that an investor has, for example, a competitor as one of its portfolio companies.

C. Assess the appropriate structure for the equity investment.

Once an appropriate equity investor has been identified, the equity investment typically will proceed to the preparation of a term sheet that identifies the key terms of the investment, as well as a diligence request and the execution of a confidentiality agreement to facilitate the exchange of information to the investor for the investor’s due diligence purposes. A term sheet is a helpful means of assessing whether the parties truly see eye-to-eye with each other on the critical aspects of the investment before expending significant time and expense negotiating definitive documents, and may include the following terms:

- The identification of the relevant entity that will receive such funds (e.g., will the investment be made into a special purpose vehicle solely created for the project (for example, a Project Company) or will the investment be made into the Sponsor which may hold assets unrelated to the Solar Project);

- The amount of the investment, as the Sponsor should ensure that it receives sufficient capital to minimize future dilutive “cram-down” financings, but also not take in more capital than is needed as this also will have dilutive effects to the existing shareholders. Milestone-based investments may help serve to mitigate Sponsor risk in terms of securing additional future financing, while helping investors stage their investment to ensure that the Sponsor can meet specific financial and commercial targets before disbursing additional funds; and

- Whether the equity security will be “common stock,” which is typically issued to founders, optionees, and “angel” investors, or “preferred stock,” which not only is senior to the common stock in preference but also typically has additional terms and conditions that increase preferred stockholders’ return on investment and control over the Sponsor such as:

• The right to appoint one or more board members;

Industrial Biotech


• Dividend rights;

• A “liquidation preference,” which is the right to receive a preferential return on investment in the event of a liquidity event such as a merger, asset sale, or change of control;

• A “redemption right,” which is the right to redeem the equity securities at an agreed-upon point in the future;

• “Anti-dilution rights,” which protect an investor from the dilutive effect of future equity issuances; and

• “Protective provisions,” which allow the preferred holders certain veto rights over key corporate actions.

D. Have candid discussions to ensure that expectations are aligned on key business issues, including:

- The market opportunity;

- The company’s ability to execute its business plan, and the investors’ commitment to both the initial and subsequent capital needs during the company’s life cycle;


Page 199 of 212

- A realistic commercialization timeline, use of proceeds, and the expected internal rate of return of the Solar Project; and

- The appropriate liquidity event, be it an acquisition or an initial public offering, and how the investor can add value to facilitate a liquidity event (e.g., by assisting in pre-public corporate governance compliance required by Sarbanes-Oxley, or introductions to key strategic partners, customers, and potential acquirers in the future).

In summary, both Sponsors and investors analyzing an equity investment should conduct a realistic assessment of the company’s capital needs, structure an investment that can add value to the company and its projects, and seek to create a mutually beneficial working relationship where expectations on key business issues between the Sponsor and the investor are aligned.

IV. Time to Project Finance

Before beginning an examination of the project financing process, it is worth noting different options for raising debt in the context of project development. The three most frequently utilized project financing structures are the syndicated loan, the issuance of project bonds through a private placement, and the issuance of “Term B loans.”

A. Syndicated Loans and Project Bonds

Currently the majority of renewable projects are financed through the syndicated commercial loan market. Syndicated loans are loans in which a group of banks each take a portion of a larger loan, and thus minimize the risk that any one individual lender making the same loan would otherwise have. Syndication is usually coordinated by an arranger bank. An alternative to the syndicated loan market is the private placement of debt through “Section 144A” offerings, which are exempt from registration with the SEC if the purchasers are “Qualified Institutional Buyers” as defined in the Securities Exchange Act of 1933. The issuer of 144A bonds could be either the Project Company or the Sponsor. Syndicated loan structures are often preferred to accessing the capital markets through 144A offerings, because capital markets investors are generally less likely to assume construction risk, and the disclosure documentation for a 144A offering is generally more extensive than that prepared in connection with syndicating a commercial loan. In addition, amounts raised through a 144A issuance are all dispersed at closing, which leads to negative carry implications. Moreover, private placements or corporate level offerings tend to be fixed rate, which, while providing certainty, removes the upside potential of floating rates that are available pursuant to commercial bank loans. On the other hand, 144A bond offerings are generally completed more quickly and inexpensively than a syndicated project loan, the covenants contained in the governing documentation may be less restrictive, and the repayment period of private placement debt offerings is generally longer.

B. Term B Loans

Several years ago, Term B loans emerged as a subset of the project lending market and were characterized by shorter terms and lower or delayed amortization, often with bullet payments due at maturity. Correspondingly, Term B loans

Industrial Biotech


carried higher risk profiles and usually were rated non-investment grade. In addition, the terms and conditions of Term B loans tended to be less onerous than traditional project debt that amortized over a longer period. As a result of the subprime lending crisis and following credit crunch, the Term B loan market all but disappeared and has yet to re-emerge. For purposes of the following discussion, due to the considerations set forth above and the lull in the Term B market, we assume that a traditional bank syndication model of project financing will be most beneficial to the Sponsor. Although the terminology may differ from transaction to transaction, the documentation for such a project financing is governed by a credit or financing agreement (“Credit Agreement”), and at a minimum will include an asset security and equity pledge agreement, a mortgage, and various Consents.

C. Loan Types

Depending on the development stage of the project, and within the project finance framework, the Sponsor may on behalf of the Project Company, seek construction loans, term loans, working capital loans, and/or a letter of credit facility. Construction loans, as the name implies, are utilized only for the period that the project is under construction. The interest rate is generally higher vis-a-vis a term loan, but more frequent drawdowns of construction loans are permitted and at the end of the construction loan availability period, the construction loan usually converts to a term loan. The term loan is characterized by a set and limited commitment or drawdown period and an extended amortization period. Term loans usually have a lower interest rate than construction loans, and may have scheduled (quarterly or otherwise) repayment dates or set amortization schedules. The conversion from a construction loan to a term loan often coincides with the definition of “Substantial Completion” or “Final Completion” under the EPC Agreement, and a failure to achieve such conversion by a certain date may cause a default under the construction loan and accelerate the debt due thereunder.


Page 200 of 212

Working capital loans, which are used primarily for ordinary course expenses such as inventory purchases, are generally sized smaller than construction or term loans and are subject to a maximum available amount tied to the value of a Project Company’s inventory and cash (often 80%). The working capital loans are usually revolving in nature, meaning that amounts borrowed can be reborrowed once they are repaid. Letters of credit are made available on the Project Company’s behalf usually for the benefit of third parties under the Project Agreements — for example, if a letter of credit is required as credit support under a PPA, an EPC Agreement, or for the provision of utility services. In many cases, draws by a third party on an outstanding letter of credit will operate to reduce the amount of working capital loan availability.

The term of a project finance loan will vary depending on the term of the principal off-take agreement. To minimize risk profile and lower borrowing costs, the loans will ideally amortize in full prior to the end of the term of the PPA. The borrowing costs of a renewable project will invariably depend on the risk profile determined by the characteristics of the project itself, in particular the lenders’ view of the likelihood that the project will default on its loans. In addition, exposure to merchant power or other off-take risk will increase borrowing costs relative to projects that have a long-term PPA, particularly one that is fixed price with take or pay terms. The reduced risks that come with long-term PPAs prevent Sponsors from taking full advantage of arbitrage opportunities that may become available in the spot market if power prices rise, as price risk is avoided for the producer. Recent trends have seen a wide swing in terms, and it is difficult to provide standard pricing terms, although as a general rule rates are higher and fees have increased, while internal credit reviews have become more stringent and less forgiving of unmitigated project risks or even minor holes or errors in Project Agreements. Typical lending fees for a project financing include the following: (i) two percent (2%) to six percent (6%) of the aggregate loan commitment as an arranging or structuring fee, (ii) one percent (1%) of the aggregate loan commitment as a syndication fee, (iii) $75,000 administrative agency fee to be paid annually, (iv) $50,000 collateral agency fee to be paid annually, and (v) facility fees to each lender in the syndicate in an amount between three-quarters of one percent (.75%) and one and one-half percent (1.5%) of each lender’s commitment. In addition, the Project Company will be required to pay the professional fees and administrative expenses of each of the lenders in evaluating the transaction, negotiating the loan documents, and providing the loan.

Despite the non-recourse nature of pure project financing, in some transactions lenders will seek guarantees for certain obligations of the Sponsor or its affiliates, either to ensure construction of the project or to ensure that the Project Company is sufficiently capitalized to meet its debt service requirements. While by no means a requirement in all transactions, under certain market conditions, a guaranteed (or limited recourse) project finance structure may be the only way to finance one or more projects or to obtain reasonably priced project debt.

Industrial Biotech


D. Security Package

1. Overview

Project finance requires the pledge of a comprehensive collateral security package to the lenders in exchange for the making of loans. The collateral security package, in the absence of recourse to the Sponsor, serves as the basis for the lenders’ securing repayment in the case of default. Specifically, all assets of the Project Company owned at the time of the loan closing, in addition to those acquired post-closing, will be pledged to the lenders until the loans are fully repaid. Included in the assets to be pledged will be all of the Project Company’s personal property, accounts, contractual rights, and intellectual property. The Project Company’s real property is pledged to the lenders in the form of a mortgage. A pledge of the equity interests in the Project Company is executed by the Holdco or any other entities that directly hold equity interests.

As part of the collateral security package, the lenders will require a Consent from some or all of the Counterparties, together with a legal opinion from counsel to each Counterparty regarding its good standing and the enforceability of the Project Agreement and Consent. The Consent negotiation process can be time consuming and even contentious, especially if the interests of the Sponsor and the Project Company on the one hand, and the Counterparty on the other, are not aligned. To complicate matters, lenders may use the process of Consent negotiation to incorporate amendments to the relevant Project Agreement, which are likely to benefit the Project Company as well as the lenders,but at which the Counterparty may balk as a renegotiation of the fundamental business agreement embodied in the Project Agreement. Even fundamental Consent terms such as the extension of cure periods for defaults for the benefit of the lenders in the event the Project Company does not cure may be viewed as an unfavorable renegotiation from the perspective of a Counterparty. In addition, some Counterparties are hesitant to enter into a contractual relationship with a large, often Wall Street-based financial institution as a putative future partner. The prospect of perceived bargaining asymmetry often complicates what may be tedious three-way negotiations between the Counterparty, the Project Company, and the lenders, with the Project Company likely playing the role of honest broker in order to facilitate prompt agreement and closure of the financing.


Page 201 of 212

2. Distribution of Project Revenues

Almost all project financed loans have what is referred to as the “project waterfall.” All revenues received by the Project Company are placed in a master project revenue account, which serves as the top of the metaphorical waterfall. As the money flows down the waterfall it is siphoned off into segregated secured accounts at each different level as described in an Accounts or Disbursement Agreement, with any funds remaining at the bottom of the waterfall being paid, assuming there are no defaults and that certain financial tests are met, to the equity owners of the Project Company. Typically, the project waterfall is structured (roughly) in a manner as described below, with most withdrawals from the waterfall occurring on a monthly or quarterly basis as appropriate:

- The first level of payment would be in an amount necessary to pay costs incurred by the Project Company (i.e., construction and/or operation and maintenance expenses depending on the project’s stage of development), including pre-approved reasonable amounts paid to the Sponsor’s affiliates under the OMA, the ASA, and the TLA;

- The second level of payment would be to the lenders to pay: (i) loan fees and expenses, (ii) interest payments, and (iii) principal payments (in this order);

- The third level of payment will be used to fill an account segregated for the purposes of paying future debt service in times of lower project revenues, although once this account has been filled to the level of the required amount no amounts will be taken out at this level;

- The fourth level of payment is often referred to as a “cash sweep” in which the lenders are repaid outstanding principal with a certain percentage of the excess cash (generally one-third or half, which increases in a default scenario) remaining after the operation of the three waterfall levels above;

- The fifth level of the waterfall may operate to fill one or more reserve accounts, often designated for future major maintenance or other purposes, but once the reserve account is filled with the required amount no amounts will be taken out at this level;

- The sixth level of the waterfall may be used to repay the holders of subordinated debt or bondholders, if applicable; and

Industrial Biotech


- The seventh level of the waterfall allows for cash remaining after amounts have been removed at the higher levels to be paid to the equity holders of the Project Company in the form of an equity distribution, assuming there are no defaults and that financial tests are met.

While every project waterfall will operate somewhat differently and many will have features unique to specific project and financing arrangements, the waterfall operation outlined above is generally standard in project financing arrangements.


Page 202 of 212

EXHIBIT 4: TYPICAL PROJECT FINANCE WATERFALL


E. Operating Restrictions

Project finance lenders place restrictions and affirmative obligations on the Project Company that will significantly impact its day-to-day operation. While many of the affirmative obligations in particular may seem like ordinary course of business operations, and the affirmative obligations and restrictions taken individually may not seem particularly onerous, on a collective basis compliance with these obligations and restrictions requires time and effort from the Sponsor’s employees. It is worth noting in connection with the time consuming nature of complying with the covenants set forth in project financing documentation that there may be certain economies of scale, particularly where the individual projects are smaller (as is usually the case with PV Solar Projects), to arranging project financing on a portfolio basis.

More specifically, project finance lenders will require that the Project Company: (i) comply with all laws and regulations, including permits, (ii) construct and operate the project in accordance with prudent industry standards, (iii) pay its debts and obligations as they become due, (iv) use proceeds received and cash flow as set forth in the financing documentation (including operation of the waterfall), (v) maintain pre-determined (and generally quite comprehensive) insurance coverage, (vi) maintain books and records in accordance with GAAP, (vii) adopt and update budgets, (viii) permit independent verification by the lenders’ representatives of performance tests, (ix) maintain in effect all Project Agreements, (x) preserve title to all assets, (xi) update the financial model, (xii) maintain the liens granted under the

Project Construction/OperatingExpenses

Fees, Interest & Scheduled Principal

Maintain Required Debt ServiceReserve LevelCash Sweep to Lenders

Maintain Required Major MaintenanceReserve Level

Payment of Subordinated Debt(if any)

Remaining amount distributed to equityholders (assuming no defaults andfinancial tests are met)

Project Revenues Account

Construction / Operating Account

Debt Payment Account

Debt Service Reserve Account

Major Maintenance Reserve Account

Subordinated Debt Account

Distribution Account

Project Revenues Received$

Project Construction/OperatingExpenses

Fees, Interest & Scheduled Principal

Maintain Required Debt ServiceReserve LevelCash Sweep to Lenders

Maintain Required Major MaintenanceReserve Level

Payment of Subordinated Debt(if any)

Remaining amount distributed to equityholders (assuming no defaults andfinancial tests are met)

Project Revenues Account

Construction / Operating Account

Debt Payment Account

Debt Service Reserve Account

Major Maintenance Reserve Account

Subordinated Debt Account

Distribution Account

Project Revenues Received$

Typical Project Finance Waterfall

Industrial Biotech


security documentation, and (xiii) enter into pre-approved hedging arrangements both for commodity inputs — for example, natural gas in the case of a CSP project or feedstock in the case of a biofuel plant — and for purposes of interest rate protection. This list is far from comprehensive in scope or detail. In addition, comprehensive reporting requirements will be set out in the Credit Agreement that obligate the Project Company to frequently provide the lenders copies of everything from construction status reports to auditor’s letters to notices of certain adverse events.

Prohibitions placed on the Project Company by the financing documentation will likely include: (i) incurring indebtedness subject to certain exceptions, (ii) incurring liens subject to certain exceptions, (iii) making investments subject to certain exceptions, (iii) changing the nature of the business, (iv) issuing equity securities, (v) disposing of assets outside of the ordinary course of business, (vi) consolidating or merging, (viii) transacting with affiliates subject to certain exceptions, (ix) opening accounts other than those secured under the financing documentation, (x) creating subsidiaries, partnerships, or joint ventures, (xi) making certain tax elections, (xii) making certain ERISA elections, (xiii) amending Project Agreements (including EPC change orders) subject to certain exceptions, (xiv) entering into additional Project Agreements, (xv) suspending or abandoning the project, (xvi) entering into hedging arrangements not approved by the lenders, (xvii) budgeting changes subject to certain tolerance bands, and (xviii) making equity


Page 203 of 212

distributions outside of the waterfall framework and unless certain criteria are met, including achieving certain cash available to debt service ratios on a historical and prospective basis (usually between 1.25:1 and 1.5:1). While this list is not comprehensive, it should again be stressed that lenders will generally tend to be sensitive to the financial interests of the project and will to some degree tailor a covenant package to the project’s expected construction and operation characteristics. It should also be noted that project finance lenders will often entertain requests for waivers of obligations set forth in the financing documentation after the closing of the loans, as they are incentivized to keep the loans performing and out of default.

F. Potential Defaults

“Event of Default” is the legal term for the circumstance that allows the project finance lenders to exercise their remedies under the financing documentation, including acceleration of the outstanding debt and foreclosure. Events of Default may include: (i) nonpayment of fees, interest, or principal due under the financing documentation (usually with a very short grace period), (ii) breach of representation or warranty made in the financing documentation (usually with a grace period if capable of being cured), (iii) non-performance of certain covenants or obligations under the financing documentation (usually with a grace period if capable of being cured), (iv) cross-defaults to other debt instruments, (v) non-appealable legal judgments rendered against the Project Company, (vi) certain events related to ERISA, (vii) bankruptcy or insolvency, (viii) default under or termination of Project Agreements, (ix) significant delays in construction schedule, (x) failure to obtain or maintain a necessary permit or government approval, (xi) unenforceability of financing documentation, (xii) certain material environmental matters, (xiii) loss of or damage to collateral, (xiv) abandonment of the project, and (xv) a change of control. Many Events of Default have cure periods, which allow the Sponsor or Project Company to take action over the course of a certain period (usually 30 days but may be less or more) to remedy the non-compliance if the Event of Default is capable of being cured; for example, a “default” under another debt instrument may be cured by paying the amount due but a final, non-appealable legal judgment against the Project Company would be incurable. In addition, during the course of negotiating the Credit Agreement it will be important for the Project Company’s representatives to qualify as many of the Event of Default provisions with materiality and “Material Adverse Effect” standards as possible, that will provide the Project Company more leeway to avoid an Event of Default and the potential loss of the project.

G. Conditions to Closing

Project financing lenders will require that a lengthy list of conditions be satisfied in order to “close” the financing and fund the loan. While many of the precedent conditions and required documents are shared with other forms of financing, it is worth mentioning certain of the conditions that constitute particularly long lead time items that must be commenced months prior to the close of the financing. Specifically, project finance lenders will generally require the delivery of the following as conditions to closing the loan: (i) a report of an independent engineer that confirms the technology employed by the project is commercially viable, the reasonableness of budgetary assumptions, the absence of serious environmental issues, compliance with all necessary permits or approvals, and that financial projections are realistic; (ii) a power or biofuels market report (if the project will have significant uncontracted off-take) setting forth expected market conditions over the course of the loan; (iii) an Environmental Site Assessment (at least a Phase I report concluding that no further such reports are necessary); (iv) an insurance report from the lenders’ insurance consultant; (v) land surveys and site descriptions; (vi) a commodity management plan (in the case of biofuels facilities); (vii) evidence that the required equity component of the project has been contributed; and (viii) copies of all third-party and government approvals and permits. The lenders will also require the delivery of the required Consents and legal opinions from significant (if not all) Counterparties, which may require significant effort to obtain.

Industrial Biotech



Page 204 of 212

As a final note, depending on the project’s funding requirements and the size of the equity contribution and project finance commitments, it may be possible to include subordinated debt in the financing package. In the case of certain renewable energy projects (e.g., biofuels production facilities), tax-exempt state bond financing may be available to close any gap between the raised and required equity in a project finance scenario. As a general rule, subordinated debt will be more expensive than the senior debt due to the subordinated lenders’ higher risk of non-payment. In almost all circumstances, the subordinated debt will need to be in place prior to finalization of the senior project debt to avoid the substantial costs that would be incurred to re-document the senior loan. If subordinated debt is employed, an Intercreditor Agreement will be negotiated between the agent for the senior lenders and the trustee or agent for the subordinated debtholders, pursuant to which the senior lenders will obtain standard terms of subordination to ensure their senior lien and payment positions vis-a-vis the subordinated lenders and any unsecured creditors in the case of any Event of Default by the Project Company, or its bankruptcy or insolvency.

V. Tax Implications

A. Key Federal Tax Incentives for Renewable Energy Projects

The U.S. government provides several tax subsidies to encourage the development of renewable energy projects. These tax benefits are currently necessary to make renewable energy projects economically competitive with energy produced from conventional sources. These tax incentives can finance as much as approximately 60 percent of the capital cost of a project. The following is a brief discussion of some of the key federal tax incentives available to developers of, and investors in, renewable energy projects.

1. Production Tax Credits

A production tax credit (“PTC”) is available for the production and sale of electricity produced from Solar Projects and other renewable sources. Other than solar, renewable sources of energy that qualify for the PTC include wind, biomass, geothermal, municipal solid waste, and hydropower (in the case of newly installed turbines). To qualify for the PTC, electricity from these sources must be produced at a facility that is placed in service before 2009 and the facility must be located in the United States.

The PTC is available for 10 years following the date the qualified facility is placed in service. The amount of the credit for each year is generally determined by multiplying the credit rate by the number of kilowatt hours of electricity produced by the taxpayer from a qualified facility and sold to an unrelated party. The credit rate is adjusted for inflation each year and varies based on the type of renewable resource (for 2008, the credit rate for Solar Projects is 2.1 cents per kilowatt hour). The amount of the PTC is reduced by as much as 50 percent to the extent the project benefits from grants, tax-exempt bonds, subsidized energy financing, or other federal credits.

2. Investment Tax Credit

Another credit available to Solar Projects and other renewable sources of energy is the investment tax credit (“ITC”), which is equal to the product of the “energy percentage” and the taxpayer’s tax basis in its “energy property” that is “placed in service” during the taxable year. “Energy property” includes, among other things, equipment that uses solar energy to generate electricity, to heat or cool (or provide hot water for use in) a structure, or to provide solar process heat. Certain fuel cell power plants that are placed in service before December 31, 2008, also qualify for the ITC. The “energy percentage” is 30 percent for solar equipment and fuel cell equipment that is placed in service before December 31, 2008. For solar equipment placed in service in 2009 and thereafter, the energy percentage is 10%, and the ITC is not available for fuel cell property placed in service after 2008. To be “placed in service,” the property must be ready for use, which generally requires that all tests have been completed, all licenses and permits have been obtained, and the project is synchronized with the transmission system and is operational.

To qualify for the ITC, the energy property must satisfy several requirements. The property must be constructed or acquired by the taxpayer, and the original use of the property must commence with the taxpayer. In addition, the property must be used within the United States, and depreciation or amortization must be allowable with respect to the property. The property must meet any performance and quality standards prescribed by the Internal Revenue Service (“IRS”) after consultation with the Department of Energy. The property must also not be public utility property or used by a tax-exempt or governmental entity. Finally, the project must not include property which is part of a facility the production from which is used to claim PTCs.

Industrial Biotech


The ITC vests 20 percent a year over five years. If the property is disposed of prior to fully vesting, the unvested portion is recaptured. For purposes of calculating the ITC, the taxpayer’s basis in the property is reduced if the property is financed in whole or in part by tax exempt bonds or subsidized energy financing. In addition, the taxpayer’s depreciable basis in the property is reduced by 50 percent of the amount of the ITC.


Page 205 of 212

3. Depreciation

Certain equipment used in Solar Projects and other renewable energy projects may qualify for a special 50 percent depreciation deduction in 2008. To qualify for this bonus depreciation, the developer must not have been committed to the investment before January 1, 2008, and the project must go into service before the end of 2008. Any basis in an alternative energy project that is not deducted pursuant to the foregoing bonus depreciation is generally depreciated over five years following the date the project is placed in service.

4. Biofuels Credits

Both excise and income tax credits are available for biofuel blenders, users, and sellers. The credits are available in varying amounts for several types of biofuels and biofuel mixtures, including ethanol ($0.51 per gallon), biodiesel ($0.50 per gallon), agri-bodiesel ($1.00 per gallon), and renewable diesel ($1.00 per gallon). Unless extended by Congress, the biofuel credits with respect to ethanol will expire on December 31, 2010, and the biofuel credits available for biodiesel, agri-biodiesel, and renewable diesel will expire on December 31, 2008.

Blenders of biofuels generally may claim the applicable mixture credit (both excise tax and income tax credits) where the mixture will be sold or used by the blender in its trade or business. With respect to biodiesel, the blender must obtain a certification from the biodiesel producer or importer which identifies the product produced or imported and the percentage of biodiesel and agri-biodiesel in the product. Producers and blenders of biofuels are required to register with the IRS. The excise tax credit for biofuels is refundable to the extent the credit exceeds the blender’s excise tax liability. The income tax credit with respect to a biofuel mixture is reduced by the amount of excise tax credits claimed with respect to such mixture.

Taxpayers may claim the applicable biofuels income tax credit if the biofuel is: (i) used by the taxpayer in a trade or business, or (ii) sold by the taxpayer at retail to a person and placed in a fuel tank of such person’s vehicle. In addition, small producers of biofuels (both ethanol and biodiesel) qualify for an income tax credit equal to 10 cents per gallon (subject to a maximum 15 million gallons per taxable year). In general, a producer of biofuels (both ethanol and biodiesel) will qualify as a small producer if its productive capacity for the relevant biofuel is less than 60 million gallons per taxable year. Certain aggregation rules apply to determine a taxpayer’s qualification for the small producer credits. The amount of the biofuels income tax credit is included in the taxpayer’s income. Unlike the biofuels excise tax credit, the income tax credit is not refundable, and any unused income tax credit may be carried back one year and carried forward 20 years.

B. State Tax Incentives

Although not the focus of this discussion, many states and local municipalities also offer tax incentives, in various forms (e.g., income tax credits, sales and use tax exemptions, property tax exemptions, and abatements), to promote the development of projects utilizing a wide variety of renewable energy sources. An example of an available state income tax credit includes a renewable energy production tax credit established by Florida to encourage the development and expansion of renewable energy facilities in that state. This annual corporate tax credit is equal to 1 cent per kilowatt hour of electricity produced and sold by the taxpayer to an unrelated party during a given tax year. For the purposes of this credit, renewable energy is defined as “electrical, mechanical, or thermal energy produced from a method that uses one or more of the following fuels or energy sources: hydrogen, biomass, solar energy, geothermal energy, wind energy, ocean energy, waste heat, or hydroelectric power.” An example of a sales and use tax exemption is the state of New York exempting the sale and installation of residential solar-energy systems from the state's sales and use taxes. The exemption applies to solar-energy systems that utilize solar radiation to produce energy designed to provide heating, cooling, hot water, and/or electricity.

C. Tax Incentive Monetization Structures

For many reasons, a developer of a Solar Project or other renewable energy project may not be able to benefit from the various tax subsidies available to the project. Various strategies have developed that allow a developer to receive value for, or “monetize,” the tax incentives the developer would not otherwise be able to utilize. These strategies generally involve an institutional investor that can benefit from the tax incentives acquiring an equity interest in the project. Two of the main strategies, the “partnership flip” and the sale-leaseback, are discussed below.

Industrial Biotech


1. Partnership Flip

In a typical partnership flip transaction, an institutional investor will form a partnership with the developer, which will own the Solar Project or other renewable energy project. The investor will receive an allocation of tax benefits and cash distributions from the partnership until the investor achieves an agreed upon after-tax return. Subject to some limitations, the investor may make its investment in the partnership over time, which effectively allows the investor to fund its investment in the partnership with reductions in future federal income tax liability.

In the initial stage of the project, the investor generally will receive a disproportionate allocation (e.g., 99 percent) of the partnership’s income or loss and any tax credits (e.g., PTC, ITC) available to the partnership. When the investor’s target return is achieved (the “flip-point”), the investor’s allocation of partnership items is reduced to a small portion (e.g., five percent).


Page 206 of 212

The partnership generally will distribute its available cash flow 100 percent to the developer until the developer recoups its cash investment in the project, and cash would thereafter be distributed 100 percent to the investor until the flip-point is reached. Following the flip-point, cash distributions would be made in accordance with partnership allocations (e.g., 95% to the developer and 5% to the investor). The developer will typically have an option, exercisable on or after the flip-point, to purchase the investor’s interest in the partnership at its then fair market value.

In late 2007, the IRS published safe harbor guidelines for wind partnership transactions under which it would treat the investor as a partner in the partnership (rather than a purchaser of tax credits) and respect the disproportionate allocation of PTCs to the investor. Transactions that do not satisfy the safe harbor will be subject to close scrutiny by the IRS. Although the safe harbor applies only to wind transactions and the allocation of PTCs, the renewable energy industry is generally following the safe-harbor guidelines in structuring partnership flip transactions for non-wind projects and for tax incentives other than the PTC.

2. Sale-Leaseback

The sale-leaseback is another structure utilized to monetize various tax incentives. Although the sale-leaseback generally may not be used to monetize PTCs, the technique has been used extensively to monetize ITCs in Solar Projects. In a typical sale-leaseback transaction involving a Solar Project, the developer will install, operate, and maintain the Solar Project, and a customer will agree to purchase the power generated from the Solar Project under a long-term PPA. The developer will incur all expenses related to the installation, operation, and maintenance of the solar equipment.

To monetize the ITC, bonus depreciation, and other tax benefits, the developer will sell the facility to an investor. The investor will lease the project back to the developer for a lease term approximating the term of the PPA, and the developer will typically use the PPA as collateral for its lease payment obligations. The developer’s revenue from the PPA is utilized to make rental payments under the lease.

The investor is considered the owner of the Solar Project for tax purposes, and it therefore claims the ITC and other tax benefits. The investor shares its tax savings with the developer in the form of reduced rents. The developer will typically have an option, exercisable at the end of the lease term, to purchase the Solar Project from the investor at its then fair market value.

For the sale-lease back structure to work, the lease must be structured as a “true lease” for tax purposes. There is an extensive body of law addressing the characterization of transactions cast in the form of a lease. In general, a lease should be respected as a true lease if the lessee does not have an option to purchase the property for an amount less than its fair market value, the lessor retains the risk that the property will decline in value (e.g., the lessor does not have the right to require the lessee to purchase the asset at a fixed price), and at the end of the lease, the leased asset is expected to have significant residual value (e.g., 20% of its original cost) and significant remaining useful life (e.g., 20% of its originally estimated useful life).

VI. Conclusion

Companies that are in the business of developing renewable energy projects confront a host of complex and inter-related commercial and legal issues that must be successfully navigated to ensure a project’s success and realize potential investor returns. Regardless of whether project finance is employed, it is important for Sponsors to assemble a team of professional advisors that can not only assist in executing a debt or equity transaction, but also analyze the various options that may exist in the course of developing projects. The currently tight credit environment, which is characterized

Industrial Biotech


by a lack of liquidity in the marketplace and a general risk aversion on the part of lenders, serves only to heighten competition for available debt. However, in such an environment, the right combination of business model, project scale, contractual structure, and equity support will still be attractive to project lenders for long-term debt commitments.

Determining whether to pursue project financing in the course of developing renewable projects is one of the most fundamental decisions that developers must make. An affirmative decision will dictate the legal and contractual structure of the projects, place certain operational limitations on how the projects operate, and limit the developer’s discretion regarding the use of much of the cash flow from the project. On the other hand, a successful project financing can maximize equity returns through increased project leverage, remove significant liabilities from the Sponsor’s balance sheet, capitalize on tax financing opportunities, and protect key Sponsor assets. In order to take full advantage of project financing opportunities, it is vital that companies invest the time and resources during the initial development stages to obtain the best possible terms and conditions in commercial agreements which serve as the foundation to project financing on successful terms.


Page 207 of 212


ANALYST CERTIFICATIONSI, Michael McNamara, certify that all of the views expressed in this research report accurately reflect my personal views aboutthe subject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be, directly orindirectly, related to the specific recommendations or views expressed in this research report.

I, Laurence Alexander, certify that all of the views expressed in this research report accurately reflect my personal views aboutthe subject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be, directly orindirectly, related to the specific recommendations or views expressed in this research report.

I, Paul Clegg, CFA, certify that all of the views expressed in this research report accurately reflect my personal views about thesubject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be, directly orindirectly, related to the specific recommendations or views expressed in this research report.

I, Alex Barnett, CFA, certify that all of the views expressed in this research report accurately reflect my personal views about thesubject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be, directly orindirectly, related to the specific recommendations or views expressed in this research report.

I, Debra Bromberg, certify that all of the views expressed in this research report accurately reflect my personal views about thesubject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be, directly orindirectly, related to the specific recommendations or views expressed in this research report.

I, Robin Campbell, Ph.D., certify that all of the views expressed in this research report accurately reflect my personal viewsabout the subject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be,directly or indirectly, related to the specific recommendations or views expressed in this research report.

I, James Harris, certify that all of the views expressed in this research report accurately reflect my personal views about thesubject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be, directly orindirectly, related to the specific recommendations or views expressed in this research report.

I, David Paek, certify that all of the views expressed in this research report accurately reflect my personal views about thesubject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be, directly orindirectly, related to the specific recommendations or views expressed in this research report.

I, Lucy Watson, certify that all of the views expressed in this research report accurately reflect my personal views about thesubject security(ies) and subject company(ies). I also certify that no part of my compensation was, is, or will be, directly orindirectly, related to the specific recommendations or views expressed in this research report.

Important Disclosures

As is the case with all Jefferies International Ltd. employees, the analyst(s) responsible for the coverage of thefinancial instruments discussed in this report receive compensation based in part on the overall performance of thefirm, including investment banking income. We seek to update our research as appropriate, but various regulationsmay prevent us from doing so. Aside from certain industry reports published on a periodic basis, the large majorityof reports are published at irregular intervals as appropriate in the analyst's judgement.

For Important Disclosure information on companies recommended in this report, please visit our website athttps://jefferies.bluematrix.com/bluematrix/JefDisclosure or call 212.284.2300.

Meanings of Jefferies International Ltd. RatingsBuy - Describes stocks that we expect to provide a total return (price appreciation plus yield) of 15% or more withina 12-month period.

Hold - Describes stocks that we expect to provide a total return (price appreciation plus yield) of plus 15% or minus10% within a 12-month period.

Underperform - Describes stocks that we expect to provide a total negative return (price appreciation plus yield) of10% or more within a 12-month period.

Our focus on mid-capitalization and growth companies implies that many of the companies we cover are typicallymore volatile than the overall stock market, which can be amplified for companies with an average stock priceconsistently below $10. For companies in this category only, the expected total return (price appreciation plusyield) for Buy rated stocks is 20% or more within a 12-month period. For Hold rated stocks with an average stockprice consistently below $10, the expected total return (price appreciation plus yield) is plus or minus 20% within a12-month period. For Underperform rated stocks with an average stock price consistently below $10, the expectedtotal return (price appreciation plus yield) is minus 20% within a 12-month period.

NR - The investment rating and price target have been temporarily suspended. Such suspensions are incompliance with applicable regulations and/or Jefferies International Ltd. policies.

CS - Coverage Suspended. Jefferies International Ltd. has suspended coverage of this company.

NC - Not covered. Jefferies International Ltd. does not cover this company.



Restricted - Describes issuers where, in conjunction with Jefferies International Ltd. engagement in certaintransactions, company policy or applicable securities regulations prohibit certain types of communications,including investment recommendations.

Monitor - Describes stocks whose company fundamentals and financials are being monitored, and for which nofinancial projections or opinions on the investment merits of the company are provided.

Valuation MethodologyJefferies International Ltd. methodology for assigning ratings may include the following: market capitalization,maturity, growth/value, volatility and expected total return over the next 12 months. The price targets are based onseveral methodologies, which may include, but are not restricted to, analyses of market risk, growth rate, revenuestream, discounted cash flow (DCF), EBITDA, EPS, cash flow (CF), free cash flow (FCF), EV/EBITDA, P/E,PE/growth, P/CF, P/FCF, premium (discount)/average group EV/EBITDA, premium (discount)/average group P/E,sum of the parts, net asset value, dividend returns, and return on equity (ROE) over the next 12 months.

Risk which may impede the achievement of our Price TargetThis report was prepared for general circulation and does not provide investment recommendations specific toindividual investors. As such, the financial instruments discussed in this report may not be suitable for all investorsand investors must make their own investment decisions based upon their specific investment objectives andfinancial situation utilizing their own financial advisors as they deem necessary. Past performance of the financialinstruments recommended in this report should not be taken as an indication or guarantee of future results. Theprice, value of, and income from, any of the financial instruments mentioned in this report can rise as well as falland may be affected by changes in economic, financial and political factors. If a financial instrument isdenominated in a currency other than the investor's home currency, a change in exchange rates may adverselyaffect the price of, value of, or income derived from the financial instrument described in this report. In addition,investors in securities such as ADRs, whose values are affected by the currency of the underlying security,effectively assume currency risk.

Distribution of RatingsIB Serv./Past 12 Mos.

Rating Count Percent Count Percent

BUY [BUY/ SB] 467 58.30 51 10.92

HOLD [HOLD] 297 37.08 22 7.41

SELL [SU/ UNPF] 37 4.62 2 5.41

OTHER DISCLOSURESThis material has been prepared by Jefferies & Company, Inc. a U.S.-registered broker-dealer, employingappropriate expertise, and in the belief that it is fair and not misleading. The information upon which this material isbased was obtained from sources believed to be reliable, but has not been independently verified, therefore, we donot guarantee its accuracy. Additional and supporting information is available upon request. This is not an offer orsolicitation of an offer to buy or sell any security or investment. Any opinion or estimates constitute our bestjudgment as of this date, and are subject to change without notice. Jefferies & Company, Inc. and JefferiesInternational Limited and their affiliates and their respective directors, officers and employees may buy or sellsecurities mentioned herein as agent or principal for their own account. Upon request Jefferies & Company, Inc.may provide specialized research products or services to certain customers focusing on the prospects for individualcovered stocks as compared to other covered stocks over varying time horizons or under differing marketconditions. While the views expressed in these situations may not always be directionally consistent with thelong-term views expressed in the analyst's published research, the analyst has a reasonable basis and anyinconsistencies can be reasonably explained.

Additional information for UK and Canadian investorsThis material has been issued by Jefferies & Company, Inc. and in the United Kingdom ("UK") is approved anddistributed by Jefferies International Limited which is authorized and regulated by the Financial Services Authority("FSA").

In the UK, this material is intended for use only by persons who have professional experience in matters relating toinvestments falling within Articles 19(5) and 49(2)(a) to (d) of the Financial Services and Markets Act 2000(Financial Promotion) Order 2005 (as amended) or to persons to whom it can be otherwise lawfully distributed. ForCanadian investors, this material is intended for use only by professional or institutional investors. None of the



investments or investment services mentioned or described herein are available to other persons or to anyone inCanada who is not a "Designated Institution" as defined by the Securities Act (Ontario). For investors in theRepublic of Singapore, this material is intended for use only by accredited, expert or institutional investors asdefined by the Securities and Futures Act and is distributed by Jefferies Singapore Limited ("JSL") which isregulated by the Monetary Authority of Singapore. Any matters arising from, or in connection with this materialshould be brought to the attention of JSL.

The information set forth herein was obtained from sources believed to be reliable, but has not been independentlyverified by JIL, JSL or JEFCO. Therefore, except for any obligation under the rules of the FSA we do not guaranteeits accuracy. Additional and supporting information is available upon request. This is not an offer or solicitation ofan offer to buy or sell any security or derivative instrument, or to make any investment. Any opinion or estimateconstitutes the preparer's best judgment as of the date of preparation, and is subject to change without notice. JIL,JSL, JEFCO, their associates or affiliates, and their respective officers, directors, and employees may have long orshort positions in, or may buy or sell any of the securities, derivative instruments or other investments mentioned ordescribed herein, either as agent or as principal for their own account.

This material does not constitute a personal recommendation or take into account the particular investmentobjectives, financial situations, or needs of individual clients. Clients should consider whether any advice orrecommendation in this report is suitable for their particular circumstances and, if appropriate, seek professionaladvice, including tax advice. The price and value of the investments referred to herein and the income from themmay fluctuate. Past performance is not a guide to future performance, future returns are not guaranteed, and a lossof original capital may occur. Fluctuations in exchange rates could have adverse effects on the value or price of, orincome derived from, certain investments.

Jefferies & Company, Inc. research reports are disseminated and available primarily electronically, and, in somecases, in printed form. Electronic research is simultaneously available to all clients. This report or any portionhereof may not be reprinted, sold or redistributed without the written consent of Jefferies & Company, Inc. JefferiesInternational Limited has adopted a conflicts management policy in connection with the preparation and publicationof research, the details of which are available upon request in writing to: The Compliance Officer, JefferiesInternational Limited, Vintners Place, 68 Upper Thames Street, London EC4V 3BJ; telephone +44 (0)20 70298000; facsimile +44 (0)20 7029 8010.

© 2008 Jefferies International Ltd.





Notes




Notes

Jefferies Research


Jefferies & Company, Inc.

Jefferies Research O

ctober 2008C

lean Tech

nolog

y Primer

Key Themes in Clean TechnologyMichael [email protected] (0) 20 7029 8680Laurence Alexander, [email protected] 284 2553Paul Clegg, [email protected] 284 2115

SolarPaul Clegg, CFAMichael McNamaraJames [email protected] (0) 20 7029 8691David [email protected] 284 2175

WindMichael McNamaraJames Harris

Biofuels Laurence Alexander, CFARobin Campbell, [email protected] (0) 20 7029 8678Lucy [email protected] 284 2290

BioplasticsRobin Campbell, Ph.D.Laurence Alexander, CFALucy Watson

WaterAlex Barnett, CFALaurence Alexander, CFALucy Watson

Carbon SequestrationLaurence Alexander, CFAMichael McNamaraPaul Clegg, CFA

Battery TechnologyAlex Barnett, [email protected] 1 5343 6714

Fuel CellsMichael McNamara

Nuclear Debra E. [email protected] 284 2452Laurence Alexander, CFA

Project Finance Laurence Alexander, CFAChris Groobey, Baker & McKenzie [email protected] 835 4240Nathan Read, Baker & McKenzie [email protected] 835 1668

102008 UK Clean Technology Primer_cvr:Layout 1 10/13/2008 9:29 AM Page 1

Member SIPC • © 10/2008 Jefferies & Company, Inc.

US Equity Sales Offices

Atlanta, Georgia 404 264 5000Boston, Massachusetts 617 342 7800Chicago, Illinois 312 750 4700Dallas, Texas 972 701 3000Houston, Texas 713 658 1100Los Angeles, California 310 445 1199New York, New York 212 284 2300San Francisco, California 415 229 1500Short Hills, New Jersey 973 912 2900Stamford, Connecticut 203 708 5900

International Equity Sales Offices

London, England +44 (0) 207 618 3500*Paris, France +33 (0) 1 5343 6700*Tokyo, Japan +81 (0) 3 5226 2411**Zurich, Switzerland +411 227 1600*

*Offices of Jefferies Group, Inc. subsidiaries Jefferies International Limited

**Jefferies Japan Limited

Jefferies US & International Sales Offices

Jefferies & Company, Inc.520 Madison AvenueNew York, NY 10022

www.jefferies.com

Investment Banking

Sales & Trading

Research

Asset Management

Economy & Finance

Clean tech industry primer jefferies (2008)